• Register
  • Help
Results 1 to 6 of 6

Topic: Which Version Of Skating Is Better?

Share/Bookmark
  1. #1

    Which Version Of Skating Is Better?

    Hi folks, I've just completed a new version of Skating. I tried to make it sound less computerized by randomizing the velocity, as well as using the Sonar Quantize plug-in to alter the start times of notes. I used these effects sparingly. It may be most noticeable on the flute solo in the middle of the piece. Is it better? Worse? Little or no difference? Any feedback would be helpful.

    Here's the link:

    http://www.mpdmusic.com/listening-area.html (There are two versions, the second version is the revised one).

    Thanks,

  2. #2

    Re: Which Version Of Skating Is Better?

    Hello, Michael!

    I've had a very nice time this morning, skating on your virtual ice rink again this morning. There are so many lovely, romantically gliding passages in this. I also think it's remarkable how many different mood shifts there are throughout the piece, and all so in keeping with your theme.

    As for comparing the two versions, the differences are so subtle as to be mostly unnoticeable. I think you've gone through what was hopefully at least a mildly interesting period of experimenting, but I know your interest is in composing, not becoming a technical wizard at musical renderings. You are in a large and fine company of composers in that regard, so I would suggest you relax and just go forward with your writing.

    But really, I would be hard pressed to notice the difference between these two recordings.

    It's extremely difficult to re-engineer a project file that's already been completed. There are edits and tweaks, such as you've done, but it just isn't the same as when the music is initially put together in a different way in the first place. Nothing can really replace using a keyboard and playing both notes and crucial controllers like CC1 or 11 for volume in Garritan Libraries. Some people get adept at compensating for the lack of a keyboard, but it really is a lot of work, and as I said above, it's not the sort of time-consuming effort that a lot of composers really want to get in to.

    Since you have done this major re-working of your files, I don't want to minimize your efforts. And since you did that work, I'll point out a few things:

    --Maybe it was just an omission, but in your text in this post, you didn't mention working with CC1/11 volume control. You mention "randomizing the velocity" and "using the Sonar Quantize plug-in to alter the start times of notes," but you didn't list volume control which is the most critical control for breathing life into the Garritan sample sets. Maybe you did some of that, but if so, perhaps it was done only tentatively. The sustained string notes at the very end, for example, seem to enter and remain at one volume for their duration, instead of swelling, or fading.

    The Quantize function rounds notes off precisely to chosen note value - That's not what to use for altering start times. The CAL routines I pointed out earlier are good for randomizing start times, but nothing beats moving notes by hand, so there's a logic and intelligence behind the edit choices being made, like pushing notes slightly ahead of the beat on brisker passages, for instance.

    Here's a screen shot of Sonar's Piano Roll View showing the use of some basic MIDI controls:



    In the control panes displaying the data below the notes: Velocity - The values aren't random, but logically used so that some notes have a fast attack, and other notes have a gentler, slower attack. It's very important to understand that velocity does not control volume in Garritan, it only effects the way notes begin. That's why when you said a lot of your editing was with the velocities, I wanted to point out that this alone can only produce subtle differences - and simple randomization can't really get the most musical results since there's a logic to when you use high or low values.

    The next pane is CC1 volume control. Notice how it's fluctuating constantly over the course of the passage. That data was played and recorded live in real time. It emulates the ebb and flow of volume which is a crucial part of any instrument's performance when it's played live, and something we can reproduce with MIDI.

    VOLUMES OF YOUR MP3s - I downloaded the first Skating so I could look at it in my audio editor. The overall volume is low, with the loudest sections only using half of the available range. When we click to hear, the music sounds weak because its scale of volume isn't at its optimum. I ran a simple Normalize plugin, and instantly the music was up to the level where it should be. The normal setting for my system's volume knob didn't need to be re-set, and the impression of your music was stronger. Of course people can turn their volume knobs as needed, but a lot of people won't do that - They click on an MP3, and if it's too soft when played at their usual volulme settings, they get a negative impression about the music. It's just the way people listen.

    ALL right - but with all that said, I want to get back to my earliest point - That you have a lovely composition here. Your primary interest is in writing, not becoming an engineer. I'm not sure you'd be happy to divert your time and attention into trying to get more out of MIDI, because no matter how you look at it, it Does take a lot of one's time. You've done a worthy experiment, tweaking your files, but it's important to remember that retro-producing existing files is extremely difficult, time consuming, and tedious. Maybe you'll want to adapt the way you work on new things, because using all these controls in the first place is Much easier than trying to change work already done.

    I hope this reply is useful. My main purpose here, actually, is applaud for you and your dreamy piece, "Skating."

    Randy

  3. #3

    Re: Which Version Of Skating Is Better?

    Delightful piece - but I cannot discern the differences and say there is noticeably an improvement in the output file as there are other possible issues in the way.

    I think there is a problem with the structure of the string arrangement that 'flattens' it too much. They don't sound like they are 'full' and always include the triad - It's hard to tell due the hollow tone from them. Perhaps they are not spaced out enough - have you attempted to include the 5 string sections, vlns 1, 2 violas, celli, bass?

    Very little dynamics at play I think. It improves on this front once the other instruments appear.
    Website:
    www.grahamplowman.com
    YouTube Music:
    My Channel
    Twitter:
    @GPComposer
    Facebook:
    Facebook

  4. #4

    Re: Which Version Of Skating Is Better?

    Quote Originally Posted by rbowser- View Post
    Hello, Michael!

    I've had a very nice time this morning, skating on your virtual ice rink again this morning. There are so many lovely, romantically gliding passages in this. I also think it's remarkable how many different mood shifts there are throughout the piece, and all so in keeping with your theme.

    As for comparing the two versions, the differences are so subtle as to be mostly unnoticeable. I think you've gone through what was hopefully at least a mildly interesting period of experimenting, but I know your interest is in composing, not becoming a technical wizard at musical renderings. You are in a large and fine company of composers in that regard, so I would suggest you relax and just go forward with your writing.

    But really, I would be hard pressed to notice the difference between these two recordings.

    It's extremely difficult to re-engineer a project file that's already been completed. There are edits and tweaks, such as you've done, but it just isn't the same as when the music is initially put together in a different way in the first place. Nothing can really replace using a keyboard and playing both notes and crucial controllers like CC1 or 11 for volume in Garritan Libraries. Some people get adept at compensating for the lack of a keyboard, but it really is a lot of work, and as I said above, it's not the sort of time-consuming effort that a lot of composers really want to get in to.

    Since you have done this major re-working of your files, I don't want to minimize your efforts. And since you did that work, I'll point out a few things:

    --Maybe it was just an omission, but in your text in this post, you didn't mention working with CC1/11 volume control. You mention "randomizing the velocity" and "using the Sonar Quantize plug-in to alter the start times of notes," but you didn't list volume control which is the most critical control for breathing life into the Garritan sample sets. Maybe you did some of that, but if so, perhaps it was done only tentatively. The sustained string notes at the very end, for example, seem to enter and remain at one volume for their duration, instead of swelling, or fading.

    The Quantize function rounds notes off precisely to chosen note value - That's not what to use for altering start times. The CAL routines I pointed out earlier are good for randomizing start times, but nothing beats moving notes by hand, so there's a logic and intelligence behind the edit choices being made, like pushing notes slightly ahead of the beat on brisker passages, for instance.

    Here's a screen shot of Sonar's Piano Roll View showing the use of some basic MIDI controls:



    In the control panes displaying the data below the notes: Velocity - The values aren't random, but logically used so that some notes have a fast attack, and other notes have a gentler, slower attack. It's very important to understand that velocity does not control volume in Garritan, it only effects the way notes begin. That's why when you said a lot of your editing was with the velocities, I wanted to point out that this alone can only produce subtle differences - and simple randomization can't really get the most musical results since there's a logic to when you use high or low values.

    The next pane is CC1 volume control. Notice how it's fluctuating constantly over the course of the passage. That data was played and recorded live in real time. It emulates the ebb and flow of volume which is a crucial part of any instrument's performance when it's played live, and something we can reproduce with MIDI.

    VOLUMES OF YOUR MP3s - I downloaded the first Skating so I could look at it in my audio editor. The overall volume is low, with the loudest sections only using half of the available range. When we click to hear, the music sounds weak because its scale of volume isn't at its optimum. I ran a simple Normalize plugin, and instantly the music was up to the level where it should be. The normal setting for my system's volume knob didn't need to be re-set, and the impression of your music was stronger. Of course people can turn their volume knobs as needed, but a lot of people won't do that - They click on an MP3, and if it's too soft when played at their usual volulme settings, they get a negative impression about the music. It's just the way people listen.

    ALL right - but with all that said, I want to get back to my earliest point - That you have a lovely composition here. Your primary interest is in writing, not becoming an engineer. I'm not sure you'd be happy to divert your time and attention into trying to get more out of MIDI, because no matter how you look at it, it Does take a lot of one's time. You've done a worthy experiment, tweaking your files, but it's important to remember that retro-producing existing files is extremely difficult, time consuming, and tedious. Maybe you'll want to adapt the way you work on new things, because using all these controls in the first place is Much easier than trying to change work already done.

    I hope this reply is useful. My main purpose here, actually, is applaud for you and your dreamy piece, "Skating."

    Randy
    Thanks Randy for some great feedback, as usual. You are right, it may not be in my DNA to produce midi masterpieces. Unless I can acquire the skills to put the feeling in up front, I don't think I'll be able to improve the sound much. I was just trying to see what some mild tweaking might do. I'm not surprised by the responses I'm getting; I can hardly hear the difference myself. As for volume, I did do a little minor tweaking. The quantize plug-in I used is the one that does alter the start times. It has a randomize function. (There are 2 quantize effects in Sonar). However, for some reason it didn't seem to work right. I don't really know either how to mix down and normalize. I have the mains volume at -7. Then I used the Vintage Channel plug-in to equalize and compress, and bring it up to a listenable level. I'll have to look at what tools I have to normalize the mp3 file. I used to use Audio Cleaning Lab, and it had a great normalizing tool in it. Maybe Sonar has something I could use?

    I really appreciate the positive feedback on the music itself. The music is, after all, what it's all about. I was never much of a performer. Defintely more the composing type. Even when I do scores, I tend to skimp on putting in dynamic markings, thinking that the conductor and performers will figure all that out. I know that's not true, and that it's the composer's job to indicate how the piece should be performed. For some reason, dynamic variation is not included in my gifts. Melody, harmony, counterpoint, form - I can do all that. But dynamics are incredibly hard for me. I'm not making excuses, I really am deficient there. Maybe I've got some damage to that part of my brain! I'm actually toying with the idea of paying someone to realize my music for me. But it would probably cost as much as outfitting myself with my own studio, considering that I've composed 5 multi-movement works. A couple of CD's worth of music.

    Well, I guess for now I will just continue to try to get my works into decent enough shape that I can put them on my website, in hopes some conductor will hear them one day, and contact me about a performance. I know it's a long shot, but we've all got to dream. ("Lose your dreams and you will lose your mind" - Rolling Stones, Ruby Tuesday).

    Appreciate all the help,

  5. #5

    Re: Which Version Of Skating Is Better?

    Quote Originally Posted by Plowking View Post
    Delightful piece - but I cannot discern the differences and say there is noticeably an improvement in the output file as there are other possible issues in the way.

    I think there is a problem with the structure of the string arrangement that 'flattens' it too much. They don't sound like they are 'full' and always include the triad - It's hard to tell due the hollow tone from them. Perhaps they are not spaced out enough - have you attempted to include the 5 string sections, vlns 1, 2 violas, celli, bass?

    Very little dynamics at play I think. It improves on this front once the other instruments appear.
    Thanks, Graham for the feedback. Strings are so difficult, aren't they! I have a strange tendency to write 5 part counterpoint with the strings, which probably does stretch them a bit thin at times. The begining of Skating has the first violins in double octaves with the violas, on the melody. The second violins are doing a little secondary melody between them. The cellos have yet another melodic line, and the bass is the bass. Very unorthadox, but that's the only way I could find to make it sound good. And yes, the dynamics are a real problem for me. I'm rather clueless what to do about it (see my response to Randy's comments).

    You're right, there's essentially no difference between the two versions. I think maybe I didn't want to change it too much, out of fear that I would only make it worse. If only I had your abilities as a midi arranger! I just was listening to your "Two Towers" realization. It's incredible. If I could ever get my music sounding that good, I could probably sell it!

    Thanks,

  6. #6

    Re: Which Version Of Skating Is Better?

    Hello again, Michael

    Quote Originally Posted by michael diemer View Post
    ...The quantize plug-in I used is the one that does alter the start times. It has a randomize function. (There are 2 quantize effects in Sonar). However, for some reason it didn't seem to work right...
    Ah, that's right. You mean the quantize function in that Step Recorder mode - That's just not in my orbit, I forget about it. It's extremely rare for me to use any quantize. But that's still applying quantization - the definition of that is rounding off the start times to a selectable note duration. I suppose if you have it on 64th or 128th notes it could help, but it's still limited. If you want to do anymore of this sort of thing again, try the Random Time CAL routine that you can apply to any amount of MIDI you select. All the CAL scripts are under CTRL+F1 by default.
    Quote Originally Posted by michael diemer View Post
    ..I don't really know either how to mix down and normalize. I have the mains volume at -7. Then I used the Vintage Channel plug-in to equalize and compress, and bring it up to a listenable level..
    -7 is very low. You need to do more than just make a subjective call about whether things are at a "listenable level," because that's subjective and not a measure against professional recording levels. One of the most common issues I see people posting at the Sonar Forum is typically under a subject line like "Why are my recordings so soft?"

    Use the compressor's output level control, inch it up while watching your Master fader's level in Sonar. If you don't use a Master fader, then watch the fader which shows your sound card/interface's level. Move things up until you start getting the red horizontal bar at the top - the peak meter. Don't worry about the vertical meter getting into the red, that's just your headroom, the extra mile you can go. As soon as that top peak warning lights up, you've gone too far. Bring it a hair under that, until in the loudest section you have no peaks. Your recordings will be much louder.

    If you mix down your audio right there in Sonar, you can then use Sonar's built in Normalize function under Audio>Effects. Just never Normalize to full 100%. I usually use just a bit over 98%. When you make an MP3 from a .wav 2-track master, the encoding boosts the volume slightly. Your MP3 will peak out even though the original doesn't.
    Quote Originally Posted by michael diemer View Post
    ...For some reason, dynamic variation is not included in my gifts...
    Being a bit flatline and cautious with dynamics seems to be rather common in MIDI composers/musicians. My theory is that their focus is so much on the notes and getting them to work together, that they're not getting the importance of then trying to emulate the instruments meant to play those notes. The only instrument I play is piano, but all the rest of the orchestra I take a stab at emulating. We've all heard what those instruments sound like, and we can always refreshen our sense memories by listening to good recordings. The hallmark trait of all instruments being played in performance is that they are positively alive with dynamics, virtually undulating with volume changes - especially when instruments are played in solo. I think if you think more in terms of the specific instruments and how they're played, rather than just the general sound of the ensemble, you would more naturally fall into being aware of and using dynamics more.

    Hang in there, Michael!

    Randy

Go Back to forum

Tags for this Thread

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
  •