• Register
  • Help
Results 1 to 6 of 6

Topic: Delay in final output

Share/Bookmark
  1. #1

    Delay in final output

    A delicate question.
    Orchestral piece with piano solist.
    I have 6 rows, seen from the listener:

    row0: solist
    row1: all strings (violins1,-2,violas, cellos,contrabasses)
    row2: flutes, piccolo, oboes
    row3: clarinets, bassoons
    row4: all brass (french horns, trumpets, trombones, tubas)
    row5: timpani and percussion toys

    all rows have their own Garritan Ambience setting, merely they only differ in pre-delay(just about 10 ms=10 feet), relation between wet and dry (1 dB). For the solist I used a different setting, a piano has its own rights....

    Sonar and also Adobe Audition have all the wave-tracks of all instruments, "bussed" into those rows mentioned. In the final rendering I notice some "timing" problem, best be heard when in a 4 beats measure(bar) the 1st and 3rd are hit by the orchestra and the 2nd and 4th by the solist at "supposed" to be even intervals. But it is not an even interval. The piano comes in earlier (just a tiny bit).

    This also occurs when the piece starts. Not all instruments start at exactly the same moment. Row 4 comes in a but earlier.

    Anybody? Why is this? The final rendering is: "export to audiofile" of all tracks. Some latency problem? Some harddisk problem? The reason why I did this with both programs is to eliminate "oh-that is Sonar" or "Audition does that every time".

    Please help,

    Raymond

  2. #2

    Re: Delay in final output

    Hopefully someone can help you, but I for one don't really understand your workflow. The questions that spring to my mind are:

    1) You have six rows - where, what tool are you using that has six rows?
    2) you have 'bussed' WAV tracks from Sonar and Audition into the rows - the rows in what? why? How? How did they get there into Sonar and Audition in the first place? If they are already WAV files in eg Audition, why are you bussing them into 'rows' in another tool?
    3) when you are rendering - where are you rendering from?

    What you are doing sounds unrealistically complex - if you have your GPO instruments running in a midi sequencer like Cubase or Sonar, you could just 'render' or 'export' from there in a one step process. There's nothing stopping you having 6 instances of Ambience running each with its own settings in Sonar simultaneously (which is what I think you want) - and render directly from Sonar in one step - that way all the tracks will stay locked to clock. You could group your instrument outputs into Mixer Groups so you have six master faders - one for each group.

    My guess is that maybe you are rendering each section with its own ambience setting, creating a WAV of that orchestral section and then trying to audio mix and finding the files aren't in sync? I have to say from my experience (years ago) that two identical files rendered by Cubase at different times of day would not actually be in sync with each other - it seems incredible, but its true - maybe that is what you are seeing? The one-step render will solve that.

    Lastly if you are married to your current workflow, to experiment you could repeat your workflow but with identical Ambience settings - as that is the only variable you are mentioning.

  3. #3

    Re: Delay in final output

    I've read that this needs some explanation.
    • In Sonar there are multiple tracks which have calls to GPO instruments.
    • Those tracks are in fact MIDI files with parts of the original score (e.g. tremolo for violins are on a separate track and calls the tremolo-violin in GPO)
    • Then I export all piccolo, all flutes, all violins, etc. to their individual wave tracks. I do this for editing reasons with Audition.
    In a separate project in Sonar I import those wave tracks again and assign a reverb to a group of instruments- in a bus, called row x. So all strings are in row 1 (the front row in an orchestral setup - more or less) and all flutes e.a. are in row 2. The rest is in the previous message.

    Somehow it is not possible on my system to have more than three Ambience plugins running in real time. I tried different settings of DMA buffering, latency, etc. but the system runs into dropout very soon. That's why I have to split my workflow into several stages.

    Once the instruments are grouped into busses/rows I listen carefully to all instruments in that group if they balance (in the mixer view in Sonar). I do this for every group/row. Then I play the whole lot (without any effects!! - dry) to see if the overall balance is right and adjust where necessary.

    The moment I am satisfied I activate the FX-settings and output/export the all-in-one wave file. Last night I found out that the timing wasn;t perfect, so I did it again this morning. Now I rendered all groups/rows separately with their individual Ambience settings and got 6 wave files to be combined in Adobe Audition to one final output (I could have done that with Sonar but I did it with Audition, works just the same). And now there isn't that timing problem!! This proves that somehow in the process (Sonar as well as Audition - see previous message) my system can't handle this

    All this is because my system can not handle all things together in one go.

    I thought I could render the wave file (combining all rows with their Ambience setting) in just one go and got that timing problem. I still don't know why. You mainly commented on my workflow.

    Raymond
    Asus P4B533 mobo, 2.26 Ghz CPU Intel
    1.5 GB DDR RAM
    2 HD systems, Seagate Barracuda > 120 Gb each IDE
    The Audio files and system/program are on different HD's
    Geforce4 64 Mb videocard
    Audiophile 2496 audiocard

  4. #4

    Re: Delay in final output

    Raymond, I think you're saying that you prefer to do some editing in Audition instead of Sonar, thus your projects are shuffled between applications.

    I believe you wouldn't have problems if you would settle in to do all your editing in Sonar. It's a very sophisticated program and should have all the editing tools you need.

    It's certainly true that many of us don't have systems which can play entire projects "in one go." But when the final 2 track mix-down is rendered, we have all the effects, instruments, edits, etc. that we intended for a piece.

    I work in Sonar, and make it a habit to bounce tracks to audio throughout a project as I go. I'm able to have at least one lead instrument for each group play together when everything is still in the MIDI realm. I have to live without the luxury of hearing this scratch version with reverb--but that's not a hardship really. Reverb settings are the icing applied when all the actual compositional and recording work has been done.

    I urge you to re-think your work flow, as Low was suggesting. Work entirely in Sonar, and don't export .wav files, but keep them inside the project by bouncing tracks. As Low said, there will never be an issue of sections staying in synch.

    When you have mixed some groups together--I end up with an average of 10 tracks for instance, then apply the reverb settings the way you want, and mix down the entire thing--even if there are clicks and hesitations on live playback, there of course won't be in the resulting 2 track mix.

    rbowser

  5. #5

    Re: Delay in final output

    Quote Originally Posted by rbowser-
    I believe you wouldn't have problems if you would settle in to do all your editing in Sonar. It's a very sophisticated program and should have all the editing tools you need.
    Including audio editing? If so I haven't found it yet. By the way, I looked thru the Sonar manual and couldn't find anything about those Sonitus plugins. Do you know where I can find some extra info about the plugins?
    It's certainly true that many of us don't have systems which can play entire projects "in one go." But when the final 2 track mix-down is rendered, we have all the effects, instruments, edits, etc. that we intended for a piece.
    And that was the problem. I mixed down the whole lot and found out in the final mix that the timing wasn't perfect.
    I work in Sonar, and make it a habit to bounce tracks to audio throughout a project as I go. I'm able to have at least one lead instrument for each group play together when everything is still in the MIDI realm.
    I'm trying that too. The piano is The Grand (Steinberg) and this instrument consumes so much CPU and memory that even when loading Sonar tells me to sell ice-creams, pizzas, and delivers an error-message.So I made a sub-project in Sonar, with that piano only. Now Sonar doesn't complain. Output to a wave file and imported into the main-Sonar-project.
    I urge you to re-think your work flow, as Low was suggesting. Work entirely in Sonar, and don't export .wav files, but keep them inside the project by bouncing tracks. As Low said, there will never be an issue of sections staying in synch.
    Did that earlier this week and give it a try after Christmas.
    When you have mixed some groups together--I end up with an average of 10 tracks for instance, then apply the reverb settings the way you want, and mix down the entire thing--even if there are clicks and hesitations on live playback, there of course won't be in the resulting 2 track mix.
    As I said before, I noticed the opposite. That's why I did the group-by-group output to wave files and mixed them in Audition, without the delay/hesitations.

    Thanks for your reply.... and I certainly will re-think my workflow again.

    Raymond

  6. #6

    Re: Delay in final output

    Hi, Raymond--I'm seeing your new addition to this thread, and am glad to hear that you're determined to establish the best work flow that you can.

    Something I wasn't responsive enough to in my previous post here, is how it is indeed helpful to have a dedicated audio editor outside of Sonar. I was urging you to keep your work inside Sonar to avoid your rather mysterious synchronization problems--but I see now that I was under estimating your need for the more detailed audio editing one can do in a dedicated program.

    I use Sound Forge in conjunction with Sonar. It's very conveniently set up as a Tool in Sonar, so at any point during work on a project, I can select an audio track and take it to Sound Forge for some work. But Sonar doesn't close. When I'm through with the audio file, I save it, close Sound Forge, and I'm instantly back in Sonar which then asks me if I want to keep the now revised version of that .wav file. I say "yes" and the "tool copy" now replaces the unedited old version of that track.

    I don't know if it's possible for Audition to become a Tool in Sonar like Sound Forge, but if so, it would completley eliminate the problem of shuffling between programs.

    From looking at your system set up, you have more Ram and a more powerful computer than me. Yet I'm having none of the timing issues that you're experiencing--so I still feel there's something overly complicated in your work process.

    From your new post:

    I mixed down the whole lot and found out in the final mix that the timing wasn't perfect.
    Try un-checking the "fast" bounce or export. I sometimes have missing sections in audio bounces if I use the default fast option. Maybe this is causing some problems. Give your system a break and just render to audio in real time--it may clear up your problems.

    Then you were talking about rendering to audio throughout a project, the way I said I do. You were talking specifically about the piano which of course demands a lot of computer resources. You said:

    So I made a sub-project in Sonar, with that piano only. Now Sonar doesn't complain. Output to a wave file and imported into the main-Sonar-project.
    This was touched on earlier on this thread, but this really isn't the way to accomplish what you want. All you need to do is highlight/select the piano's MIDI track and VSTi audio track, then choose "bounce" not "export." I suggest, as I did above, that you un-check the "fast" option--render it in real time, but the resulting .wav file will be right there in your project on a new track. There is NO need to "export" and then re-import the audio, or to make a sub-project of just the piano. Exporting is for when you are archiving a final 2-track master mix, or needing to get an audio file into another app for some reason.

    I hope you take that last paragraph to heart. You are making things Much more complicated for yourself than you need, and you're inviting glitchy issues like your non-synchronization problem.

    In the last part of your post, you talked again about exporting section by section, re-importing, in order to avoid the timing problems. I repeat the above advice, un-check "fast" when you bounce or export. Try it first INSIDE the project. If you want to do some editing on the resulting audio files in Audition, then hopefully you can set up that program as a "tool"--the way I described the way I use Sound Forge, to avoid all this exporting and re-importing.

    Best wishes.

    rbowser

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
  •