I certainly didn't mean to say anyone here was mindless, I just meant a general "mindless" application of crossfade is sometimes done when it shouldn't be used at all.  In the example of 16ths and 8ths you should never use crossfade. It is pointless, as note-on velocity works perfectly for short notes.  It is only sustains and legato where one might want to use crossfade. 

But an example of avoiding crossfade and using dynamic samples and note-on velocity is using a crescendo  sample followed by a leap up a fifth, played by overlapping a dynamic sample with an ff legato at the moment the dynamic reaches ff.  It is true that sort of thing won't always work depending on timing and the line.  Another example is using slight decrescendo samples instead of sustains and connecting the end of one of those to a legato p.  That can be very effective in a line which has slightly more pronounced attacks on each sustained note that then tail off to a softer legato.   

 I have a question about this. So this means you can hear some sort of phasing when this happens? What about the SE users, they'll get a lot of phasing too, with solo instruments. I know to use dynamic instruments but those are only in extend versions. So that more of a realistic performance for solo instruments you need to get full solo libraries (to be able to use dynamics)?

Cheers,

Ricardo

One reason you may want to use modulation for staccato dynamics is if you want to use velocity to trigger long/short variations. 

 That is true but one thing I've noticed is crossfaded short notes never sound as good as note-on velocity combined with amplitude which is what you get in a normally programmed multi-velocity patch.  For example a staccato will have four actual recorded dynamics, and the rest of the variation is amplitude controlled by velocity within each of those samples.  It is close to perfect for almost any line. But when you layer a little of an mp with a little of an mf sample it sounds thicker, not as clean and natural sounding, and is quite audible. And if there is no reason to do it...  though you are right about that being a good way to control it especially if you are playing live.

I've often had occasions when the change in timbre from one velocity sample to the one above is so pronounced that it causes difficulty - particularly on mf legato melodies with longish notes. I would have thought the best solution would be an increase in the number of velocity levels for the solo instruments. Eight would be a good number - obviously extend the recording time but as they would be solo rather than ensemble maybe not too expensive.

Also on the theme of dynamics be great to have some pfp percussion samples - timp swells anyone? They can be fudged but not as good as an actual recording.

ACtually on timpani I've found that using crossfade works perfectly especially if you use two patches: one for single notes and one for rolls.  This allows you to do an accented start to a roll - a classic timpani effect  such as FFp followed by a crescendo - that sounds exactly like live, or a release of a roll to a single hit that is accented more than the release sample would be - another thing often heard in timpani parts.  The one cell for single hits is set for no cossfade, the other for rolls is set for crossfade. 

I should be putting sound examples of these up ! 

In my opinion, I don't think that VSL team needs to fix this matter but I know that the crossfade is a cheap and a "cheat" and it is a way for expressive legato and sustain phrases. This matter for The Special Edition users is not ok but with the right velocities it can sound good. Any questions, of how this problem should be fixed correctly for legato and sustain phrases on some instruments? I do know that staccato phrases, crossfade is no good and individually you write the velocities for each note.

Cheers,

Ricardo

I am sorry to put a damper to your enthusiasm, but don't you think there is a reason why the big collections cost considerably more than the special editions?

But however, the number of layers for the legato patches is in most cases the same with the full instruments. Have you seriously thougt through your suggestion to record more velocity layers? For perf_legatos? Amount-of-samples-to-be-recorded-wise this is like recording (and editing) half of the instrument again. This might be considered by VSL for future projects (for instance, the Dimension Strings perf_leg has already 4 instead of the usual 2 layers) but surely not something that will happen as an update soon!

But, if there are certain instruments that cause noticeable more problems than others, like the oboe, a re-programming, so that the range where actually two samples are heard is smaller, could indeed be a solution.

Massmover is completely right on this but he is being polite.  This post is totally incorrect:

"In my opinion, I think VSL team needs to fix this matter (soon). I know crossfade is cheap and a "cheat" but it is the only way for expressive legato and sustain phrases(if there is another way, please give me a reply)..."

There is no matter to be fixed.  It has been stated that crossfade works perfectly for ensembles, and for solo one can use either crossfade or, in other cases, the dynamic samples.  All of this is planned carefully to deal with the various issues that come up with dynamics.  The post suggesting it is a matter not being dealt with is way out of line and based on simple ignorance of what is actually recorded in the samples and how it can be used.  This person needs to learn more about the library before saying something needs to be fixed.

O.K. -  goodbye you wonderful people,  I am out of here.  I tried to post some helpful info but it is, apparently, pointless. 

Thank you both Massover and William, but before going further... I am not an Ignorant. I never paid any attention to these topics in the first place, but like every learner and every student. It is my righteous duty to study and gain knowledge about it. Even if I made an "Incorrect" statement... does It mean I have to be an ignorant and arrogant? If you call me an ignorant or any other name, I don't really mind actually. Respect needs to be present for everyone around. What William said, I'll take this into a deep study and gain more knowledge about VSL samples and how it works(crossfading and other things). I do have patience of what you wrote and even if you criticize me the most, I'll just ask for feedback, instead of posting wrong statements or making false accusements.

Thank you for your feedback.

@ricardojvc6 said:

...I am deeply sorry about it.

You are OK Ricardo

The sentence above clearly shows it.

Nevertheless, I try to give you a feedback for getting more the understanding of the situation.

I am 54 and I believe William "can count a similar amount of years". So we are another generation. When we had your age we learned:

- "Why do you believe that you are right - with no experience?

- "be modest"

- "thinking before talking" 

... and lots of such things.

Young people today come with an "unblieveble and hugely" self-convidence in our eyes - so that we (the oldies) feel some statements very impolite or cheeky.

Try to understand this:

Why do you believe, that you - young, inexperienced, as a beginner - are legitimated to say...

• that a lot of developmend engineers and
• a product that has been released years ago
• thousands of professional musicians

didn't seen that here is a fault which needs to be fixed... ?

Another example

Not long ago a student wanted to make a period of practical instruction in my company. Though he never earned money before he promised me that my profit will increase after his time beeing with me...

Why did he believe that he can increase my profit? He only saw schoolhouses... This appears arrogant in the eyes of us 50+

I agree you are not an ignorant or anything similar. But we "oldies" really can't understand this huge self-convidence of young people. We called that a bluff when we have been young. On the other hand I see that this seems to be the current standard today all over the world. So you are doing nothing wrong so far. Your reaction seems to be the modern one.

Nevertheless, I hope you also can understand why a statement of a beginner (...need to be fixed) appears like a provocation for our generation.

One last thing

William and I are not only "oldies" in life we also are old hands here in the forum (See when we joined here).

Unfortunately we are a bit tired of all those beginners who start to use VSL-samples, then it doesn't work how they believe it sould and the next step is a post here "there is a bug to fix".

So please also understand William's goodby and our tiredness in this point. The story with the bugs here and there over the last 10 years get on our nerves sometimes.

Nevertheless, we are still willing to help and to share our huge amount of collected experiences.

OK, dear Ricardo

I tried to show the other side of your point of view. It isn't important to fix which one could be the right one. The most important thing is the understanding of the other side.

I wish you a lot of success in using samples of VSL.

All the wonderfull demos will show you, that a lot of things are possible.

And if not: There are still real players around who like to play just the music you want...

Beat

 Thank you Beat Kaufmann. 

I do understand, very much. If I have any question, I might post them in the forum where I can have all the help of the other users. I understand your point of view completely. I don't want to seem rather helpless so I tried to help but It didn't seem to be any good to me; It indeed made me look cheeky and arrogant. I don't want to be a person that knows everything. I just don't want to be a bad person because of what happened.  

By the way, I saw your VI Pro Tips&Tricks and help and I find them to be very useful and helpful for me to understand more about the world of Samples. You wrote before "Try to play with the samples till you find the one you like the most". It seems pretty reasonable for me and I understand that completely. I find the dynamics to be very useful for me, most of the time when I want very nice crescendos and diminuendos. I will see more about your VI Pro Tips&Tricks and VSL website questions. It really can be very helpful for someone who is new to the world of samples! 

Thank you very much!

Cheers,

Ricardo

@William said:

But an example of avoiding crossfade and using dynamic samples and note-on velocity is using a crescendo  sample followed by a leap up a fifth, played by overlapping a dynamic sample with an ff legato at the moment the dynamic reaches ff.  It is true that sort of thing won't always work depending on timing and the line.  Another example is using slight decrescendo samples instead of sustains and connecting the end of one of those to a legato p.  That can be very effective in a line which has slightly more pronounced attacks on each sustained note that then tail off to a softer legato.   

 Well I was going to sign off from this but since you are asking me directly and say "I'd love to hear an audio file..."  first of all you can hear LOTS of audio files I've done right here on this website though they may not be specific enough for you. Secondly--- 

I would need to see the actual musical passage you are talking about and not some vague description like what you gave.  Short legato - what exactly are you talking about?  Let's see a notated musical passage and then I'll get back to you.  Maybe.  But if it takes too much time, you'll have to pay to get another response.  [H] 

Hello Wrathy

Thanks for the scor excerpt.

About the oboe

It depends on how it shall sound in the end. Should you hear some accents within the 8 notes? Should there be no accents even not on beat one.

I understand your wish "legato" that you dont want any gap between the notes.

3 possibilities (without X-Velocity)

Possibility 1

Use the repetition samples. Chose those which are so long (itself) that no gaps can happenbetween notes.

In other words: The single repetition sample should sound longer than it sounds within the group.

Result: It will sound in legato-style but not so how a musician would play it. So...

Infact, it isn't that easy...

Possibility 2

I prepared the sequence from bar 125 till 129.

Unfortunately I didn't know anything about the tempo and about accents on 1 and on 3 ...?.

Nevertheless, I made an interpretation which will show you 3 things:

• 1. Use a lot of different samples for reaching the musical aim.
• 2. Use samples not because of their names but because of their sounds (already mentioned in a post above)
  
• 3. Use a DAW for adjusting volumes from sample to sample, for inserting controller curves and other things.

See: Your sequence

See: Image of the rolleditor

Listen: Played with Oboe 2

What about the samples for the first four 1/8s:

I played the first note with a single note (portato-medium) because the legato-samples often don't start with enough attack with their starting note.

Then I changed to the legato-articulation with the 2nd note, 3 and 4 I played with repetitions. So that's the basic recipe I used.

Of course I always repeated this combination for the next four 1/8 and so on (...more or less). I always used samples that sound best. Therefor I used a diminuendo instead the portamento-medium...

Bar 128: It would be also possible to play each 4 notes with legato and start new with the next 4 notes. If the starting notes sound too weak at 128 with just legato you could use a sample with more attack (fp ?) for them.

Observe the roll editor of the example (Image above):

I used several samples and I had incredibly to adjust the volumes (Velocity) that all the samples fit to each other. This leads to a next point:

If you like to get out all of the possibilities of the samples you need to use a DAW.

It doesn't make sense to mark sfzs, diminuendos etc within a score just for getting the right sound with the samples.

You also can't set every second note a p, an f, an mp and so on.

So if you really want to have a nice interpretation of your music then compose it, export a midifile, import it in a DAW and "make music then".

Be happy that you are able to listen to a quite nice orchestra with your score program but do not expect more.

Best

Beat

PS

Seeing all the tasks you need for just some notes: Producing 3minutes of music means also having a job for 3 days and more.

That's good Beat !  

I too actually experimented with this line but am too lazy to put it up on the internet.  Also, I would have to put it on a separate page on my website and no way am I doing that (unless -  as I mentioned before - you pay me - then I'm all over it !!!).  But anyway I did this keyswitching -  legato, followed by staccato, (with freshly attacked first notes of each 4-note phrase -  i.e.,  bee-uh-b-b) and then just sustain-legato followed by a short dynamic sample crescendo at the very end.  Using all note-on velocity control, no crossfade, with a little CC11 expression to enhance the dim./cresc. 

This particular example would never be good with crossfade though.  Because you would hear the phasing.  No amount of recording by VSL or any other library will improve that.   It has to be done with velocity and dynamic samples, even if it isn't easy.   And yes, 16 to 32 velocity layers would be nice, but you can't get everything you want yet.  Though VSL is working on it, I realize.

One thing to keep in mind is that velocity consists of two things: actual timbral changes in the samples, and simple amplitude changes within each range of timbral.  So that accomplishes most of what is needed, though there are tricky lines that need more work and the dynamic samples are needed for them.

BEAT!

Thanks so much for your efforts and detailed reply. Very instructive.

This is what I currently have:

Link: Excerpt from the piece

Copy this if the above attempt at direct linking does not work (copy and paste into browser).

http://www.andrewrathbun.com/mp3s/WWQuintet_excerpt.mp3

I'm on the road for a few weeks away from my rig, but will try your way after I return, to see if it yields more realism. This is going to be played by real people but I felt like taking the time to delve into VSL for some better mock up chops. And since you mentioned it, I write with pencil and paper, then, if mocking up, I record the piece in Digital Performer, then copy in Finale. Usually I just go right to Finale, but in this case I felt like mocking it up.

I really appreciate the time you took to create such a through and detailed answer

best.

I listened to the excerpt and actually thought that oboe part you have sounds very good, not at all artificial-sounding !  Though the notation has a crescendo marking to the end of the phrase and if that is intentional, it could have a 2 second crescendo sample slipped in at the very end.  Often going from legato to sustain or dynamic can be done and still sound like part of a legato phrase.

@William said:

I listened to the excerpt and actually thought that oboe part you have sounds very good, not at all artificial-sounding !  Though the notation has a crescendo marking to the end of the phrase and if that is intentional, it could have a 2 second crescendo sample slipped in at the very end.  Often going from legato to sustain or dynamic can be done and still sound like part of a legato phrase.

Thanks William. I'll definitely give that a go when I get back to town. Thanks for the input.