Mix Bus assignable to LR Bus

Forums Forums SQ Forums SQ feature suggestions Mix Bus assignable to LR Bus

Tagged: 

This topic contains 49 replies, has 18 voices, and was last updated by Profile photo of ThisIsAnAudioAccount ThisIsAnAudioAccount 7 months ago.

Viewing 15 posts - 31 through 45 (of 50 total)
  • Author
    Posts
  • #97063
    Profile photo of nottooloud
    nottooloud
    Participant

    @keithjah

    Auxes are auxiliary busses, i.e. supplementary to the main LR mix.
    Therefore you should not be able to assign these to the LR mix as they are separate by design.

    If auxes could go to the mains, I wouldn’t have to preconfigure auxes vs groups. A post-fader aux with all faders set at 0 or off would be functionally identical to a group, except it would have more flexibility.

    I’m used to building groups out of auxes on boards that didn’t have groups. The number of times I’ve switched sends-on-fader to a group on my GLD and wondered why the faders didn’t move is not a small number.

    #97070
    Profile photo of Mfk0815
    Mfk0815
    Participant

    For me this „auxes cannot be routed to main“ rule has no architectural or design reason but is more a matter of the ideology of A&H and their idea of how mixing have do be done. To change their mindset could be a hard work, I am afraid.

    #97073
    Profile photo of KeithJ A&H
    KeithJ A&H
    Moderator

    @mfk0815 – It’s not an arbitrary decision, nor do we ever intend to tell any engineer how to mix, though I appreciate the architecture can influence this quite a bit. Whilst just a picture, the diagram at the bottom of this article ( https://support.allen-heath.com/Knowledgebase/Article/View/sq-signal-path ) shows how the processing for the LR and Aux channels occurs in parallel, and Aux -> LR would require the signal to go ‘backward’ or more accurately the whole signal path to be extended. So it is this way by design and is one of the reasons for the extremely low latency and coherency of the console.
    For example, if we were to have Aux -> LR, then you could end up with the following signal paths:

    A) Input > LR (<0.7ms)
    B) Input > Group > LR (<0.7ms, coherent/sample accurate with A)
    C) Input > Group > Aux > LR (>0.7ms)

    So in the case of C, do you either ignore the coherency and potentially introduce cancellation/comb filtering, or compensate by adding latency to every other channel?
    The former has the potential to ruin a mix without obvious reason to the engineer, whilst the latter reduces the performance of the console, and then do you stop at Aux to LR or is the next request for other mix to mix routing further compounding the issues?

    Our mindset is only that we want every engineer using our console to be able to work quickly, avoid pitfalls (especially in the heat of a show) and ultimately get outstanding audio results. We are very open to discussion, but would never implement something that doesn’t improve things for everyone.

    Cheers,
    Keith.

    #97076
    Profile photo of Mfk0815
    Mfk0815
    Participant

    Thank you for that reply, Keith. So, some questions came in my mind, when I read it.
    First, why is it possibleto assign an aux to a channel on the dLive? And second, whyis it not possible to do that on the SQ? Third, there are already situations where the coherence of different paths to the Main bus can be corrupted by the user, think about inserting effects either from the FX rack or, even worse, external ones. The engineer needs to know that to avoid comb filters and so on. And the engineer also have to know the difference off DEEP plugins and FX Rack plugins.
    For me there are two different major aspects regarding mixbusses. First how to bring signals to the bus. Main differences are pre fader, post fader and subgroup style, where a subgroup can be seen as a post fader bus with fixed level at 0 dB. The second aspect is what to do with the summing signal. If i know that
    A) Input -> Main
    B) Input -> Bus(Aux/Subgroup) -> Main
    Would be aligned as but
    C) Input -> Subgroup -> Aux -> Main
    Is not aligned that would be the same type of knowledge as the stuff with insert effects and DEEP Plugins for me.

    Just my 2 cents.

    #97078
    Profile photo of Mfk0815
    Mfk0815
    Participant

    One last question regarding this issue, Keith
    What is the difference regarding latency between
    A) Input > LR (<0.7ms)
    B) Input > Group > LR (<0.7ms, coherent/sample accurate with A)
    C) Input > Group > Aux > LR (>0.7ms)
    And
    A) Input > LR > Matrix
    B) Input > Group > Matrix
    C) Input > Group > Aux > Matrix
    D) Input > Group > LR > Matrix

    Because the second sample is possible with the SQ. When do we start to be not phase coherent?

    #97080
    Profile photo of KeithJ A&H
    KeithJ A&H
    Moderator

    @mfk0815 – Great questions!

    While the processing algorithms in SQ are the same as those found in dLive they are different systems with different capabilities for different users.
    Mix freedoms in dLive do allow you to use a mix as an external input for another, and you can also do some other funky routing like input to matrix and returns to input channels.
    Some of this does cause confusion even for dLive super users though and I’m not sure most SQ users would (or should) need to be as careful as you obviously would be with the system knowledge you have.
    Referring back to the OP for example, parallel compression using an aux would be the absolute worst case scenario for coherency and so Aux -> LR is not the solution.

    DEEP processing does not add and latency of course, so there is no need to worry about that, and RackFX are most commonly used as Mix -> Return (no coherency concern) or inserted on an individual channel or mix. With inserted FX, there is the wet/dry mix (which can be used without issue), so the only way this might cause problems is if double patching an input and inserting FX on one of the channels. With this setup I’d suggest it’s more likely to have one ‘dry’ and one ‘wet’ version so you would still not be blending two non-aligned signals. (But you’re right, there is some potential for messing things up, it just takes some effort!).

    Back on the argument for Aux -> LR, the main question I see is:
    “What are the use cases for Aux -> LR which cannot be achieved through other means and which outweigh the potential for less experienced users to experience problems?”

    Thanks,
    Keith.

    PS – just saw the follow up question, which is explained here https://support.allen-heath.com/Knowledgebase/Article/View/phase-coherent-mixing-in-sq

    #97085
    Profile photo of nottooloud
    nottooloud
    Participant

    @keithjah

    That’s a clear explanation of your reasoning, thanks. I can’t think of a use case I would have that would be worth blowing coherency.

    #97087
    Profile photo of Mfk0815
    Mfk0815
    Participant

    @keithjah
    Frankly spoken, I have no obvious use case in mind for routing auxes to MainLR at the moment. That is also true when I work with the dLive, and there I have the option.
    And I know that even if I want to do that, I can somehow do a workaround.
    Generally spoken is one of the big advantage of digital consoles the possible flexibility of signal processing. Everybody who ever did a customization of the tap points on an analog console like a MixWiz or GL knows the pain to do that. On the digital console it is just a simple adjustment. The more flexibility I have the more I will use it and I will adjust my workflow step by step. „The appetite comes with the eating“ is something we say in my corner of the world.

    #99128
    Profile photo of Dilettant
    Dilettant
    Participant

    PLASSE stop compare 96KHz versus 48KHz architectures consoles…

    Sorry, i have to state this:

    No, at least myself, but i could bet many other people, too, definitely won’t.

    I simply see no reason to care about having 48 or 96 kHz on the Console core. Thats just one of many technical Parameters which may influence things I care about, but then i will care on them and not on sampling frequency.

    In Fact, at least my personal goal and i think the goal of most SQ Users is to mix Audio for humans, not for bats, dogs or any kind of analyzing device. Humans are “pets” that can hear a maximum of around 20 kHz, maybe 24 kHz as a Baby, but as far as i know there was never detected a human on the whole planet who can hear higher Frequencies.

    Most of us use Microphones which are limited to a maximum of 12 up to 20 kHz or synthesizers that use 44 or 48 kHz Sampling Frequency on our Inputs. And most of us generate output that will be shipped to “End Users” with that sampling rates.

    So why should I not compare Consoles with 48, 96, 72, 80 or 192 kHz? They all more or less can do what i want. The Sampling Frequency is the way, not the target. In Himalaya Climbing where the way may be the target, in Audio Processing it is definitely not, at least not for me.

    Of course, I may care about number of Inputs, features, noise, sound, latencies and many other things. I also may care about file sizes, bandwidth or storage, about a reasonable User Interface, about robustness of the Hardware and last but not least about money.

    If SQ is better or worse in some of that points because of using 96 kHz, that’s good, but it makes no difference if another Console would do that with better Software, other Hardware or whatever else may help. In both Cases the result is mainly the same.

    So don’t be disappointed there, I and many others will continue comparing Mixers without caring at all about what sampling Frequency they use. We also may compare Cars without caring about the inside color of the gear housing.

    #99129
    Profile photo of volounteer
    volounteer
    Participant

    @Dilettant

    You do not understand enough about digital and what happens with processing.

    96 is better than 48, especially if you end up with CDs.
    And 192 is better than 96, and 384 is better than 192.
    I vaguely recall that once a high end studio even had 786 for their use.

    Now if all you are going to do is play live via the loudspeaker then it really does not matter.
    But if you record and later process many tracks in a DAW to mix and tweak then the higher resolution is much better to have.

    This is far beyond inside color or gear housing this is more like gear ratios and horsepower in the motor and those things do matter to most car buyers.

    #99130
    Profile photo of SteffenR
    SteffenR
    Participant

    higher sampling rates have audible influence in the hearing area as well…

    If SQ is better or worse in some of that points because of using 96 kHz, that’s good, but it makes no difference if another Console would do that with better Software, other Hardware or whatever else may help. In both Cases the result is mainly the same.

    sometimes the results differ just because of the sampling rate…
    some downsides of the 44,1/48kHz world are not “fixable” with additional or different software (and mostly this software uses upsampled signals to avoid the problems)

    #99132
    Profile photo of Brian
    Brian
    Participant

    96kHz processes samples twice as fast as 48kHz consoles. So generally speaking, latency times will be quicker (shorter) with a faster console.

    Also, while I am a little fuzzy on the specifics myself, watching some of the greats in the industry like Robert Scovill, it is clear that the higher sample rate consoles are better able to handle the higher frequencies. It has something to do with the math, but the higher rate consoles are better able to produce sounds 15kHz and above than lower sample rate consoles.

    #99133
    Profile photo of Mike C
    Mike C
    Participant

    96kHz processes samples twice as fast as 48kHz consoles. So generally speaking, latency times will be quicker (shorter) with a faster console.

    Also, while I am a little fuzzy on the specifics myself, watching some of the greats in the industry like Robert Scovill, it is clear that the higher sample rate consoles are better able to handle the higher frequencies. It has something to do with the math, but the higher rate consoles are better able to produce sounds 15kHz and above than lower sample rate consoles.

    Yes the faster the sampling rate the latency is shorter/faster.

    All digital processing has a fairly steep low pass ” Nyquist Filter” at half the sampling frequency, with lower sampling frequencies and depending on the design of the filter artifacts/side effects of the filter could creep into the audio spectrum.

    #99150
    Profile photo of Dilettant
    Dilettant
    Participant

    You do not understand enough about digital and what happens with processing.

    I understand that things very well. But for many users (and me including), they don’t really matter. It is simply fact that not
    everyone that buys a SQ does that because of the (at least postulated) very low latency or finbe high frequency responses. Some of them
    even use 48 kHz AR/AB Stageboxes and are perfectly happy with that. Some even have 16 kHz Low Pass filters or Shelves on their Master Bus.
    They simply care about about what they want to produce, not about sampling Frequencies.

    Most users compare Mixers due to Criteria that matters for them. That even may not be any sound parameter
    but the housing color, the weight, the preference of their dealer or their favourite brand. If that matters for them,
    it is what they should care about and it is what they should compare.

    But also the technical argument is way less clear than you insist.

    There are a lot of experienced and professional Engineers out there that clearly say 96 kHz does _not_ make anything better most of the time
    because there is simply no signal source and no signal target that can make use of the additional Data and the
    noise floor off the algorithms is way low enough for most purposes anyway. And they have points. Just try to name 5 common
    Samplers that deliver more than 48 kHz or 5 common stage Microphones that can be used over 20 kHz – you will see that is
    not easy. Even some professional Amps and speaker Management systems have a steep 22 kHz Low Pass on their inputs.

    If you detect a special 96 kHz Console sounds better than a special 48 kHz one, that might or might not be due to the
    sampling frequency. That is, because it is practically nearly impossible to make such a comparison without psychological Biases
    and because there are many other things influencing what you hear – Many “differences” reported out there are for sure due to
    some different calibrations of used D/A-Converter (even in different Modes for 48 and 96 kHz that may be different), the following
    analogue Amp stages and of course the Loudspeakers/Headphones (that are always the by far weakest part of the Chain when it comes to exact
    reproduction).

    Only one thing is absolutely clear: 96 kHz means to process twice the data volume of 48 kHz. That _may_ reduce number of
    available Channels (like it does even on SQ Drive Multitrack Recording), it may make the system to respond less fast,
    eat DSP power, consume disk space and so on. That might make the Console more expensive (it not necessarily has to,
    that is mainly a question of FPGA Market price structures and politics at the end of the day).

    A&H says, 96 kHz help reducing latency. That can matter in some situations, it cannot in some other. If you filter
    20 kHz, the latency Argument can be correct. If you filter 100 hz, it cannot because your period is 100 ms and there is
    absolutely no way to extrapolate periods longer than 4 times of the available Data that makes sense in any form.

    You say, 96 kHz even reduces processing noise. That may be but there are rare cases that noise does really matter. Most of the
    time, you have 100 and more times higher noise levels from your preamps, microphones and even ambience. And even processing noise may
    be reduced as well by dedicated oversampling instead of generally higher sampling rates for many use cases.

    And that statement also was mentioned in the forum earlier: There were tons of really good Audio material produced with
    less than 96 kHz over decades, many of it absolutely rocking the planet. If any result doesn’t, that nearly sure is not
    due to the sampling frequency of the mixer.

    #99151
    Profile photo of Dilettant
    Dilettant
    Participant

    sometimes the results differ just because of the sampling rate…

    That may be (i think much less often than most 96 kHz users postulate), but the point is, that there simply do need not care about /why/ it differs for the User.

    Thus, it _is_ completely possible, perfectly senseful and legal to compare 96 kHz with 48 kHz Systems. There is absolutely no argument to not do that.

Viewing 15 posts - 31 through 45 (of 50 total)

You must be logged in to reply to this topic.