Expected latency when using SQ as an audio interface

Forums Forums SQ Forums SQ general discussions Expected latency when using SQ as an audio interface

This topic contains 14 replies, has 6 voices, and was last updated by Profile photo of nottooloud nottooloud 2 years, 10 months ago.

Viewing 15 posts - 1 through 15 (of 15 total)
  • Author
    Posts
  • #78691
    Profile photo of Troubleshooter
    Troubleshooter
    Participant

    Hi,

    A bit qurious what latency to expect when using USB->PC recording on SQ when using SQ as a audio interface.

    Thinking of getting rid of the RME FF800+Ferrofish in my setup when buying SQ-6, I know RME has been known for
    low latancy and stability over the years, but don’t know much on the USB connection on A&H QU/SQ serie.

    //Kent

    #80781
    Profile photo of moebius
    moebius
    Participant

    bump. Very interested as well.

    #80784
    Profile photo of jdugger
    jdugger
    Participant

    Does it matter?

    Low latency is really important when you are round-tripping audio through the computer, but the whole reason I went with a digital mixer instead of a traditional audio interface is doing things like cue mixes no longer require it!

    #80805
    Profile photo of andiabwaerts
    andiabwaerts
    Participant

    Until you want to do punch-ins…

    #80824
    Profile photo of jdugger
    jdugger
    Participant

    > Until you want to do punch-ins…

    I guess I’m confused on that one as well.

    In the old days with tape and limited tracks I would worry about performing a punch well. Now I just do another take and edit.

    To answer the original question, it looks like 1.67ms is the lowest possible setting. I’m running mine at 2048 (25ms!) because it literally just makes no difference in the workflow I use what latency I have since all of my cue mixes are run through the SQ6. The computer has become basically a fancy tape machine during tracking.

    #80888
    Profile photo of Troubleshooter
    Troubleshooter
    Participant

    Well I have finally pull the trigger on an SQ6, should arrive next week, so let’s see what I end up with, I’m happy if I can run it stable at around 5-7ms…

    #80902
    Profile photo of moebius
    moebius
    Participant

    For a mostly unidirectional workflow, latency is no a big issue.

    However, having a DAW and SQ as the core of the studio with mixed ITB/OTB instruments and effects… plus back-and-forth conversions… things will add up.

    In my planned setup, I think about SQ feeding the DAW via Dante. I also run two more machines, one laptop for portable work and another computer as VSTifx host. Now there is the option to somehow get that laptop on Dante, but there is no single great solution. Or I could simply hook the laptop to the SQ and use it as the laptops interface. That´s where the latency is compared to say, an RME Digiface. THe complete chain of signal may be: MIDI controller to laptop (vsti) to SQ to Dante to outboard fx to Dante to main DAW (and maybe back to SQ-6 to speaker). And at some point your routing choices will be limited due to overall latency. It´s much nicer to get really low values, so can route more with less concerns.

    1.67 ms sounds very good to me, does anyone know if that´s at 96/24 with 32 samples ASIO latency?

    #80914
    Profile photo of jdugger
    jdugger
    Participant

    I would think Dante is a much worse choice for latency when dealing with round-trip issues like in-line plugins and software instruments.

    I can look what the lowest settings are when I get home, but I think the very shortest possible setting was 16 samples at 24/96. The board has a tiny bit of latency itself as well in addition to the USB.

    #80959
    Profile photo of moebius
    moebius
    Participant

    Comparing Dante to ADAT or MADI, I didn´t find Dante to be slower. The cable to program latency is 1ms in 1 ms out, so 2ms overall at 96/24/64 in Ableton Live for example. My RME ADAT cards and USB interface are not faster. Of course you´d need to add a bit of latency for the digital transport, but that´s with all digital protocols and we´re talking 0.25ms using PCIeR cards on both ends. So what other options would you look at, that make Dante a bad choice? I´m still building the new studio, so I need to make choices and probably didn´t consider all options.

    #101928
    Profile photo of Jline
    Jline
    Participant

    Reviving an old thread. I’m seriously considering making my SQ-5 the centerpiece of my studio. I really like how easy it is to route audio and to control multiple headphone mixes via sends on fader. The thing I haven’t figured out is how to handle overdubbing/punching in. I have always monitored through the computer and currently have an audio interface with very low latency. I am aware that you can switch back and forth from local inputs to USB inputs via the library on the SQ, to switch from live tracking to playback, but how are people handling complex overdub sessions if the latency is a bit too high such as over USB for example? I don’t think it would be practical to constantly switch from Local to USB in this type of situation. How much better is the latency when comparing Dante to USB? Thanks….

    #101954
    Profile photo of nottooloud
    nottooloud
    Participant

    As long as you are monitoring your live sources directly through the mixer, your DAW should compensate for the latency in playback sources. It’s a little trickier if you’re playing software instruments in the DAW.

    #101982
    Profile photo of Jline
    Jline
    Participant

    As long as you are monitoring your live sources directly through the mixer, your DAW should compensate for the latency in playback sources. It’s a little trickier if you’re playing software instruments in the DAW.

    Hi, thanks for the reply. I totally get what you’re saying regarding monitoring the live sources directly through the mixer. However, when the artist needs to hear their previously recorded track, you really need to switch the input of that track to USB, that is where the latency starts to become a problem. When monitoring through the DAW, it’s super easy to do overdubs if your latency is low enough. I suppose I could create a new track for the overdub, but that might get tricky to manage with multiple overdubs. Another thought is to simply send the stereo output of the DAW to the artist, but only during overdubs. This way they would hear their existing track but the source would still be monitored live through the desk. One thing necessary for this to work is to disable input monitoring when record arming a channel, otherwise, the artist would hear themselves twice, live source with no latency, and the delayed source through the DAW with latency. That would be major comb filtering!

    #101988
    Profile photo of nottooloud
    nottooloud
    Participant

    Another thought is to simply send the stereo output of the DAW to the artist, but only during overdubs. This way they would hear their existing track but the source would still be monitored live through the desk.

    That is what I would consider the normal procedure. I never monitor inputs in the DAW.

    #102035
    Profile photo of Jline
    Jline
    Participant

    That is what I would consider the normal procedure. I never monitor inputs in the DAW

    Good to know, I have never worked that way before and always monitored through the DAW with a low latency Apogee interface via input monitoring. Just curious, how do you handle punch in’s and overdubs during a recording session? Imagine you are working through a song section by section where you need to hear your previous takes, or, you have a singer that is struggling to get it right and is doing a lot of takes. What is your workflow in this situation? Thanks!

    #102041
    Profile photo of nottooloud
    nottooloud
    Participant

    I’m in Reaper, in case that matters. I mute the section that needs to be replaced, and record however many takes in an adjacent track. No tricky punch-in timing, the singer can start singing whenever they want. Sometimes I loop a section including the part to be replaced and let recording run. When they’ve got it, I select the good take, trim that to match the muted bad section, and slide it into the main track.

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

You must be logged in to reply to this topic.