• Register
  • Help
Results 1 to 10 of 10

Topic: VSL and GS & sequencer volume/velocity levels

Share/Bookmark
  1. #1

    VSL and GS & sequencer volume/velocity levels

    I recently bought some Vienna Symphonic sample libraries, and understand that it's important to let them play at their recorded levels. I'm therefore trying to figure out the correct volume and velocity defaults for both my sequencer (Cubase SX.3) and sampler (GS2.5), so that the VSL samples will play as intended. However, there are a few different volume and velocity faders/controllers in GS, as well as ones within SX, so I'm not sure which are relevant (I know I should get to know GS better!!). Can someone help me out?

    (I'm using Cubase SX.3 and GigaStudio 2.5 on the same computer - pentium 4, 2.8 GHz, with 2G Ram - and always launch SX from GS. So far, I'm using GS for MIDI only - ie, I'm not doing any audio there - as far as I'm aware!)

    GS has the following volume faders/controllers:
    Midi Mixer fader (per channel, per Port) - default is "0"; Inputs, DSP Station - default is "-6"; Aux busses, DSP Station - default is "-6"; Masters, DSP Station - default "0".

    GS also has its sample audition 'virtual keyboard' set to a default velocity of 127.

    Then, SX of course has volume AND velocity faders/controllers for each track.
    It seems to me that these over-ride my GS settings, but I haven't experimented to see what would result were I to change my GS settings -- so I'm not sure about this.

    Last, my midi keyboard has, of course, its own volume fader.

    Question 1: Which are the levels that are directly relevant to sample performance?

    Question 2: What would be the ideal "starting levels" for each of these faders/controllers?

    Many thanks, in advance, for any tips you care to share!

  2. #2

    Re: VSL and GS & sequencer volume/velocity levels

    i don't think there is a right or wrong answer to all this. i can tell you how i did it and perhaps others can weigh in with their experiences.

    i use a template that has panning and positioning as well as volume set in the GS mixer. that way all of my sequencer levels are set to zero and i know that if the balances are off, it is probably my orchestration rather than my engineering that needs work. to get the levels in GS i did a bunch of mock ups, usually just excerpts from all different periods and compared my work to recordings. they were real sloppy and i would never play them for anyone, but i gave a good sense of where everything "should" be. it was a lot of work and took some time, but the end result now is there isn't any need to mix much if at all and most of my time is spent directly on writing. hope this helps.

  3. #3
    Senior Member Bruce A. Richardson's Avatar
    Join Date
    Sep 1999
    Location
    Dallas, Texas
    Posts
    5,755

    Re: VSL and GS & sequencer volume/velocity levels

    Hi Michael,

    Actually, I would say that it is NOT important to allow the VSL samples to play back at their originally recorded levels. Each section was recorded in isolation, with relatively equal treatment, which means that the recorded levels are not particularly tied-in to a baseline balance.

    What you get with the VSL instruments is a well-sized stereo image recorded at optimum levels with very similar distance characteristics. So, you would mix using the same sorts of methodologies you'd use with individual overdub sessions in a non-sampled production. Stereo pan to your liking, utilize EQ and reverb to create your soundstage and distance modeling, and in short, mix as if you'd done a studio project in a typical nice room, one section at a time.

  4. #4

    Re: VSL and GS & sequencer volume/velocity levels

    Quote Originally Posted by Martin Bayless
    i use a template that has panning and positioning as well as volume set in the GS mixer. that way all of my sequencer levels are set to zero and i know that if the balances are off, it is probably my orchestration rather than my engineering that needs work. to get the levels in GS i did a bunch of mock ups, usually just excerpts from all different periods and compared my work to recordings. they were real sloppy and i would never play them for anyone, but i gave a good sense of where everything "should" be. it was a lot of work and took some time, but the end result now is there isn't any need to mix much if at all and most of my time is spent directly on writing. hope this helps.
    Thanks, Martin -- creating a template that can be re-used seems like a very good idea. However, I have some questions:

    1. Which levels do you set in GS -- just the Midi Mixer volume?
    2. "to get the levels in GS i did a bunch of mock ups": This seems to imply that you set the Midi Mixers to different volumes (depending on the instrument's place in the orchestra? On recorded levels of different samples?) Is that correct? Is there any "standard" volume that VSL samples seem to require (ie, at which they were recorded)?
    3. What about velocity levels, both in GS (although the only velocity I see there is in relation to the "audition" virtual keyboard - in Midi Control Surface - which, on my GS at least, seems to have a default of 127), and sequencer?
    4. Besides using panning and volume, how can one set "position"?

    I've just begun to check out my VSL libraries, and am surprised by how same-instrument samples sometimes don't match up, either dynamically or in terms of timbre. (Some samples - like Opus 1's sustained oboe, with vibrato - are really disappointing.)

    I'm also trying to figure out if the Performance Tool won't work in step-by-step sequencing (my guess is that it doesn't), and if one can only use its Legato mode in combination with other tracks/articulations if one's prepared to use the Ghost key-switch (which means loads of hassle when editing/changing articulations after the fact).

    I'm hoping these are just the frustrations of initial steps, and that I'll eventually be as impressed by VSL as everyone else seems to be.

  5. #5

    Re: VSL and GS & sequencer volume/velocity levels

    Quote Originally Posted by Bruce A. Richardson
    Hi Michael,

    Actually, I would say that it is NOT important to allow the VSL samples to play back at their originally recorded levels. Each section was recorded in isolation, with relatively equal treatment, which means that the recorded levels are not particularly tied-in to a baseline balance.

    What you get with the VSL instruments is a well-sized stereo image recorded at optimum levels with very similar distance characteristics. So, you would mix using the same sorts of methodologies you'd use with individual overdub sessions in a non-sampled production. Stereo pan to your liking, utilize EQ and reverb to create your soundstage and distance modeling, and in short, mix as if you'd done a studio project in a typical nice room, one section at a time.
    Thank-you, Bruce: I've a feeling that you're giving me some very important advice here; unfortunately, I'm so inexperienced in recording techniques, I only understand some of it. I'm trying to imagine what overdubbing a real performer would involve: same room, same mics, similar levels, but variations in performed dynamics? Ie, tweak the ins and outs so they blend with previous stuff? (Any techs you'd care to share would be gratefully welcomed!)

    Distance modelling via EQ and reverb: I recall reading something (somewhere) about harmonics/EQ and "loudness" -- that loudly played acoustic instruments will have more powerful high harmonics, that one can imitate via EQ -- but I've no idea how EQ and spatial distance relate. Care to go a bit further?

    The only text I have on recording techniques is Modern Recording Techniques, by Huber - which I've only begun. There's a lot to cover, but he seems to cover it with only very brief summaries. Any other texts you'd recommend?

    Thanks again for so kindly lending a hand!

    - Michael

  6. #6

    Re: VSL and GS & sequencer volume/velocity levels

    yes michael, i made the volume and panning adjustments in the GS mixer. as for the mock ups, no. i used the same template for everything and then made minor changes until i got something that worked very closely for everything in general. its not an exact science but with careful reprogramming of to reflect what real instruments do (louder, more brilliant in their higher registers, quieter in the lower, etc.) its possible to really dial things in. in doing so, some of the instruments' anomalies can be fixed - which is why i think the VIs, which will only allow superficial editing, are a step backwards in quality. the thing about vsl is it is almost an instrument unto itself with a learning curve similar to other instruments.

    as for velocity levels, i set them in GS edit. i don't know of a way to do that at the sequencer level unless you mean after notes are entered. in that case, yes, i do that but usually just to correct a mistake rather than compensate for the samples (which i've theoretically already evened out). as for front to rear positioning, it is generally done by adding more reverb to instruments that sit farther from the listener (e.g., percussion vs. strings). there's more to it and some great ideas in the mixing section of the vsl forum. things like considering how an sound loses bass and treble, or how the stereo field narrows the farther away it gets.

    i think the broader answer to your question is yes, you can expect it to get better over time as you move thru the learning curve. just hang in there and enjoy!

  7. #7

    Re: VSL and GS & sequencer volume/velocity levels

    Quote Originally Posted by Martin Bayless
    yes michael, i made the volume and panning adjustments in the GS mixer. as for the mock ups, no. i used the same template for everything and then made minor changes until i got something that worked very closely for everything in general.
    Thanks, Martin! I think that I had perhaps misunderstood some of your first reply: as I now understand it, the volumes you used in the "template" you created were adjusted so that the instruments would sound close to the way they would in a live orchestral situation -- eg, woodwinds softer than brass, etc. Same with panning, reverb and EQ. Once set up, acoustic imbalances in any particular score would reflect the orchestration choices rather than the 'acoustic setup'. Do I understand you correctly now? It makes a lot of sense, and I can see how helpful it could be, beyond simply being a quickly available template.
    Quote Originally Posted by Martin Bayless
    its not an exact science but with careful reprogramming of to reflect what real instruments do (louder, more brilliant in their higher registers, quieter in the lower, etc.) its possible to really dial things in.
    I'm wondering about the logistics of this: in order to make the higher registers louder and more brilliant, your template would have to use a different channel not only for each articulation/sample, but also for each range used by each articulation/sample. Is this really what you do? Or do you perhaps take a VSL sample, edit it in GS so that its harmonic profile and velocity better reflect how it would really sound in a higher register, and then save the result as a new sample? (I'm just thinking theoretically here - as I'm sure you can tell - since I've had almost no experience with any of these procedures.)

    Quote Originally Posted by Martin Bayless
    the thing about vsl is it is almost an instrument unto itself with a learning curve similar to other instruments.
    Ah, but where to find a good teacher?


    Quote Originally Posted by Martin Bayless
    as for velocity levels, i set them in GS edit. i don't know of a way to do that at the sequencer level unless you mean after notes are entered.
    I was thinking of something like Cubase's "Velocity Shift" feature: changing it to a higher value really changes any sample's level and attack. (Of course - from what I've read - it probably also distorts the sample's sound, as well.)

    Quote Originally Posted by Martin Bayless
    as for front to rear positioning, it is generally done by adding more reverb to instruments that sit farther from the listener (e.g., percussion vs. strings). there's more to it and some great ideas in the mixing section of the vsl forum. things like considering how a sound loses bass and treble, or how the stereo field narrows the farther away it gets.
    Thank-you very much for sharing all of these things with me -- I appreciate it!

    Quote Originally Posted by Martin Bayless
    i think the broader answer to your question is yes, you can expect it to get better over time as you move thru the learning curve. just hang in there and enjoy!
    Many thanks for the encouragement! You and Bruce have been very generous and kind.

    - Michael

  8. #8

    Re: VSL and GS & sequencer volume/velocity levels

    Quote Originally Posted by Bruce A. Richardson
    . . .So, you would mix using the same sorts of methodologies you'd use with individual overdub sessions in a non-sampled production. Stereo pan to your liking, utilize EQ and reverb to create your soundstage and distance modeling, and in short, mix as if you'd done a studio project in a typical nice room, one section at a time.
    Hi again, Bruce. I just re-read your reply, and realize that I'd misunderstood it the first time (I did that with Martin's, too!): I'm so preoccupied with the question of how to seamlessly join notes from different samples/tracks that are to belong to the same individual instrument, that I took your description of "overdub sessions in a non-sampled production" as referring to a "punching in" kind of re-recording! Now I see that you're referring - as you actually wrote, very clearly - to a complete recording project, in which different instruments/voices are recorded at different times - that is, overdubbed.

    If I understand you correctly, now, you're saying that since VSL's samples were ". . . recorded in isolation, with relatively equal treatment. . . [and]recorded at optimum levels ", they must be "mixed as if you'd done a studio project in a typical nice room, one section at a time", in order to make any acoustic and musical sense.

    This makes perfect sense to me, but I'm a little confused, as VSL's tutorials (by Beat Kaufmann) seem to stress the opposite. Given my track record in misinterpreting comments here, though, I'd better re-study those tutorials -- it's quite possible I got them wrong, too!


    Sorry for getting it all mixed up!

    Thanks again!

    - Michael

  9. #9
    Senior Member Bruce A. Richardson's Avatar
    Join Date
    Sep 1999
    Location
    Dallas, Texas
    Posts
    5,755

    Re: VSL and GS & sequencer volume/velocity levels

    Quote Originally Posted by michael_maberly
    This makes perfect sense to me, but I'm a little confused, as VSL's tutorials (by Beat Kaufmann) seem to stress the opposite. Given my track record in misinterpreting comments here, though, I'd better re-study those tutorials -- it's quite possible I got them wrong, too!
    Or that he did...

    What I have observed is that the instruments are recorded optimally per session, which means that a marimba, for instance, is going to play much louder than its "normal" role in an orchestra compared to, say, the trumpet section...even if both are the same MIDI and output volume.

    So, I'd be very confused that anyone would say any different, because that is definitely what I see here in output levels between sections and instruments. There is almost no way that they could have done it any other way...if you recorded everything at a "trumpet section fortissimo" level, then things like ocean drums and harps, and other softer output instruments would have been recorded so low as to be flirting with lower resolution and increasing noise floor.

    At any rate, the ear's the ultimate guide anyway. It is really difficult to even speculate on what is a "normal" orchestral level per section, since different instrumentations and different production styles will present radically different soundstage designs. That is really the best benfit of VSL, that one can so radically change the end result.

    Best regards,
    Bruce

  10. #10
    Senior Member
    Join Date
    Dec 2003
    Location
    Budleigh Salterton
    Posts
    1,477

    Re: VSL and GS & sequencer volume/velocity levels

    Quote Originally Posted by michael_maberly
    This makes perfect sense to me, but I'm a little confused, as VSL's tutorials (by Beat Kaufmann) seem to stress the opposite.
    If I was seemingly just getting started with samples of any description - I would learn and adhere to all those BK tutorials - and then break them. AFAIC, there are NO rules, except what your ears tell you. If it sounds reasonably musical at the end, there's a good chance it is.

Go Back to forum

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •