Forum Replies Created

Viewing 15 posts - 1 through 15 (of 155 total)
  • Author
    Posts
  • #85413
    Profile photo of DavidCo
    DavidCo
    Participant

    I mistook your speaker to be at a lectern, but maybe introducing one could help keep them put somewhere useful? Maybe you could use head mounted mics? Other than that, yes the human factor is your main obstacle. I would address it as such. If they are unwilling, I would simply walk away from this one. This still doesn’t make your dynamics solution any more feasible.

    DC

    #85406
    Profile photo of DavidCo
    DavidCo
    Participant

    ***content removed – personal comment that does not contribute to the thread*** You have a quiet speaker, barely speaking louder than the HVAC. I don’t see how you’re going to fix that using full band dynamics, be it upward, downward or back-to-front compression or whatever. You either need to:

    • make the HVAC quieter
    • make the speaker speak louder
    • move the mic closer, or
    • find a way of cancelling out the HVAC.

    On the last point, have you considered using a differential microphone setup? Get two similar mics (ideally a matched pair), mount them as coincidentally as possible, aim one at the speaker’s face, the other lets say at his chest. Send those mics to a couple of input channels, and invert the phase of one. When they sum, anything common to both mics (like the HVAC) will be cancelled out. As I said, use physics.

    DC

    #85377
    Profile photo of DavidCo
    DavidCo
    Participant

    Red light comes on at +15dB (-3dBFS). Clipping occurs at +18dB (0dBFS)

    DC

    #85376
    Profile photo of DavidCo
    DavidCo
    Participant

    I’ll say again: You need to fix this with physics, not DSP.

    DC

    #85253
    Profile photo of DavidCo
    DavidCo
    Participant

    A good explanation for those interested in the four basic forms of dynamics: https://www.izotope.com/en/blog/mixing/expanding-on-compression-3-overlooked-techniques-for-improving-dynamic-range.html

    My 2p: I don’t think this problem can be solved purely in the desk using dynamics. Some form of noise rejection is needed. Perhaps even look at a differential pair? Good old fashioned physics might be the solution here.

    DC

    #84157
    Profile photo of DavidCo
    DavidCo
    Participant

    What would it replace in the signal chain, the compressor?

    DC

    #84098
    Profile photo of DavidCo
    DavidCo
    Participant

    What cards do you have in which slots in the DX32?

    DC

    #84021
    Profile photo of DavidCo
    DavidCo
    Participant

    Who dangled what carrot?

    DC

    #83689
    Profile photo of DavidCo
    DavidCo
    Participant

    It’s much easier than the summing that happens when you PFL a DCA, because the audio doesn’t need to be summed as it would be for a PFL (summing the audio is a more complex function). Only the channel levels need to be added, which are available for all channels at all times (see the meter page). Just simple addition of the numbers that are the AFL levels of the channels, that’s it.

    Let’s suppose there are two input channels, both fed a sine wave from the signal generator, and we polarity invert one of them. Now let’s feed them both to an audio group but also set them as members of a DCA.

    Q: With the audio group summing the two signals together and measuring the level post sum, what does the group metering say?
    Answer: they cancel out, there is no visible signal on the meters.

    Q: If you sum the metering data of each of the DCA’s members, what would the “DCA metering” say?
    Answer: Each member’s metering data doesn’t take into account its phase or polarity, so the levels will just be blindly added together without respecting coherency, and will behave the same regardless of phase/polarity with respect to each other.

    I am a programmer, and integer sums are probably one of the simplest operations a computer can do.

    Being a programmer doesn’t qualify you to talk DSP though.. Quid in the post, please.

    DC

    #83658
    Profile photo of DavidCo
    DavidCo
    Participant

    Mr.X: Read my post on the previous page. If I had a quid for every time someone said in an unqualified statement that something was “easy to implement”, I’d be a rich man.

    I’m not sure what metering the “hottest” signal would tell the user, the level doesn’t mean anything useful. However I agree the peak LED is a totally different consideration. It could be analogous to the layer keys peaking, alerting the user something is wrong and requires investigation.

    DC

    #83030
    Profile photo of DavidCo
    DavidCo
    Participant

    How would the SQ know what time it is?

    DC

    #82419
    Profile photo of DavidCo
    DavidCo
    Participant

    What firmware version?

    DC

    #82107
    Profile photo of DavidCo
    DavidCo
    Participant

    If I had things my way, digital mixers wouldn’t have GEQ at all as they’re a hangover of the analogue days. The only advantage a GEQ has that I can think of is the band count, and if you’re utilising most of a GEQ then you’re probably doing something wrong.

    ..but that’s just my opinion.

    DC

    #81819
    Profile photo of DavidCo
    DavidCo
    Participant

    The Qu is normalised to -18dBFS. The OBS metering is in dBFS, so -18 dB on that meter should correspond to 0 dB on your Qu.

    DC

    #81798
    Profile photo of DavidCo
    DavidCo
    Participant

    If it works as intended, it isn’t a bug.

    DC

Viewing 15 posts - 1 through 15 (of 155 total)