PA Questions

[soundgals]

Hi Peter, as you know I have a NOS1 USB which I can hook up to several systems, since I own a High-End Store in Monaco.
So I have a few questions regarding using PA in different situations.

1/ Active speakers - Is it worth trying PA with these or is it likely they are already nicely Phase aligned? If it's worth trying, how to measure DC offset in these systems, where the NOS1 is directly connected to the speakers via Ballanced XLR connectors?

2/ Semi active speakers. Here the power amps feed the mid-range and tweeter drivers and the signal is also passed from those power amps to active subwoofers. I can check the DC on the main power amps; but how can I know I'm also safe with the sub-woofers?

3/ What about using a pre-amp with PA engaged? I understand PA works in conjunction with the volume control in XXHighEnd; but what if you have several sources in addition to the NOS1 feeding the same power amps? For those sources you would need volume control at the pre-amp. I thought of two approaches: a) If PA is only engaged when the volume in XXHighEnd is at the -9.9db mark, set the volume permanently at 10db and engage PA, then control the rest of the volume range via the pre-amp. Would this still allow PA to work properly or does it rely on all changes in volume having to be made by XXHighEnd? b) If the pre-amp has ht pass through or a tape loop that can provide full output, use the NOS1 through this bypass connection and the other sources via the normal connection. Presumably there would be no problems here as all volume adjustments would be made via XXHighEnd for the NOS1; but via the pre-amp for the other sources. Any additional considerations for DC here?

Thanks for your patience in answering these questions.

Geoff

soundgalleries.com

[SeVeReD]

I was wondering this type of question too because I have active subwoofers working only the nether regions.  I've checked the amps on the main system, but how to check the internally active external subwoofers.  I have a feeling that there is no real way to check for this... maybe calling the manufacture and asking.  But like a fool I've gone ahead and engaged PA.

Question?  If the subs do pass DC, would I only lose the subs eventually, or would this "bounce/bleed" back to the rest of the system, or, no way to tell?  I'm assumming this "destruction" doesn't just happen right away if I make a mistake by activating PA... or if destruction doesn't happen right away does it mean I'm in the clear?

In any case, oh well, for sure using PA at our own risk.

[PeterSt]

The quick answer : Connect your subs through the outputs of the amps. When you have the choice (LS cable inputs on the subs) that is better anyway IMO.

The not-so-easy answer : try to reach the terminals of the driver in the sub. But usually the Class-D amp in a sub can be detached (screwed out) from the back panel. From there the "LS wires" will go to the driver. So you might be able to grab the connection points (output from the PCB) from there too ...

Peter

[SeVeReD]

The quick answer : Connect your subs through the outputs of the amps. When you have the choice (LS cable inputs on the subs) that is better anyway IMO.

Oh, I do this anyway so I guess that protects it.

[PeterSt]

100% !
If there's no DC at the output of the amps of course ...

[SeVeReD]

100% !
If there's no DC at the output of the amps of course ...

Well...
There's a little bit (less than 10mv...) but it jumps up when I press stop.  Right now I'm messing with special mode, 1024 sample rate, and PA intensity 1.  It sounds awesome... untill I push stop and I get a fairly loud and disturbing crack,,, I've moved fade out to 8 and it still comes thorough the speaker. sigh ... maybe fade out 9? hehe

[PeterSt]

Dave,

I can't guarantee that it helps, but by now you should really try this all with 0.9z-7-4. So, things sure changed throughout and that includes the "stopping" and how the fade-out is applied. But point is merely :

It should not happen at all. However, there is one issue at this moment (and I must add that to the Release Notes), which is :

The fade-out happens over those rounds of "last buffer playing". So, this is 100% related to this fade-out setting, because actually it implies the times that buffer is played. But that doesn't work at all ...
I spent a whole day at solving it, but so far I couldn't; this is because what is applied to the data in the buffer a sort of backfires when in the next round; it as to incorporate what has been applied already and this is "math" I couldn't put together. It means it has to be setup differently. BUT :

Just *because* I found this actually new culprit, I suddenly can also recognize what people complain about. Look :

untill I push stop and I get a fairly loud and disturbing crack

So, there we have an example of it;
What you will be perceiving is that a 3rd (!) round of the buffer is applied wrong data. But the solution is so easy, and actually should have been there in your settings - if I only could make clear better what to do. But simply put : your buffer is too short. So, extend the buffer so much that it will never need a 2nd round and the thing won't happen for 100% sure. So, a first thing to do wrong is your mentioned "1024 sample rate" of which I suppose you mean Device Buffer Size" which won't fit in any Q1 (of 30) unless you also have Q1Factor higher than 1. So, I *know* that with a Device Buffer Size of 4096 Q1 must be at 14 to be safe (and not ~3 which would be a comparing setting to your Device Buffer Size of 1024).
So, whether you like the sound of a higher Q1 or not, when you set your Q1 to 14 and Q1Factor to 4 and leave all further the same, you won't have any cracks at the end. Now, theoretically your "crack" comprises of four passes (three I think) of wrongly connected buffer data.

Hope this helps. It should ...
Peter

[SeVeReD]

Will try -4 tomorrow. goodnight

oh
and will read again and do your post tomorrow too.

[PeterSt]

Hi there Geoff. I didn't even realize that you never posted in here until I saw your "1". Ok, welcome to an experienced "user" !

1/ Active speakers - Is it worth trying PA with these or is it likely they are already nicely Phase aligned? If it's worth trying, how to measure DC offset in these systems, where the NOS1 is directly connected to the speakers via Ballanced XLR connectors?

There won't be any differences with other speakers, which, of course, all have their own degree of "alignment". So, inherently wrong is just that. But, inherently OK is still not OK; that's our fun ...
Anyway, key point here is : it will be hard to check for DC. So, when not possible, please better forget about these.

I know, in your situation - and of course you are not the only one - it will be tough already not to forget about checking a new amp etc. first, or to just deactivate Phase Alignment at trying a new amp before checking it for DC. However, what you can do with other DACs than the NOS1 is checking its output for DC and when it is not there it is blocked on the outputs. So, OK. With the NOS1 this is not necessary because nothing is blocked.

2/ Semi active speakers. Here the power amps feed the mid-range and tweeter drivers and the signal is also passed from those power amps to active subwoofers. I can check the DC on the main power amps; but how can I know I'm also safe with the sub-woofers?

What I already said in the other post : connect the sub(s) through the outputs of the amp(s). And if they are not made for that, you can still try it with expected good results I think, because it is only a matter of being able to attennuate enough on the subs. And, I'd say this always can be done, and it's only that the small range left to adjust the sub's output level to the normal speakers could be a. rough thus b. harder to do, c. be more subject to changes because of temperature changes or even vibrations.

3/ What about using a pre-amp with PA engaged? I understand PA works in conjunction with the volume control in XXHighEnd; but what if you have several sources in addition to the NOS1 feeding the same power amps? For those sources you would need volume control at the pre-amp.

Phase Alignment indeed works in conjunction with the digital volume of XXHighEnd, but the relation of PA to the volume setting is completely lineair. A bit tough to explain what that means in this case, but let's say that the "PA net effect" is made to be the same at each digital volume setting. This means that no matter to what the volume in XXHighEnd is set, PA works out as intended. And this thus also means that as soon as it is set to -9dBFS or further down (towards -144) you can analoguely attenuate further down the line what you want. It is a digital process with analogue output and the relation is fixed in the analogue output.
One thing though :

Phase Alignment sprung from working in-DAC. But, I prefer to have it working in-amp as well. This latter is out of my control, and e.g. when the amp blocks DC at the inputs, that can't work. Now don't get confused, because where we all the time in here talk about that the amp outputs should not show DC to protect the speakers, we practically talk about blocking DC at the inputs. Still this is not the necessary case for many amps, as DC can be removed by many more means, up to at the outputs (though the latter will not be much normal). Anyway ... what this means is that when you attenuate with a preamp or other means outside of XXHighEnd, there sure can be a difference on the result compared to let do XXHighEnd all (and amps at full gain).
So, just saying, and normally DC will be blocked at the inputs of the amp.

I hope the above answers your further literal questions I did not quote.

Regards,
Peter


PS: When you have some spare time, please enter your "data" in your signature like all do. To have your website in there is okay. But it is for others to see how to interpret your remarks in the context of settings, what DAC and maybe more that you consider important to interpret you (unknown) future remarks. Thanks ...

[soundgals]

Thank you Peter. You've answered all my questions nicely. I will update my signature as you've requested

[SeVeReD]

Dave,
... But simply put : your buffer is too short. So, extend the buffer so much that it will never need a 2nd round and the thing won't happen for 100% sure. So, a first thing to do wrong is your mentioned "1024 sample rate" of which I suppose you mean Device Buffer Size" which won't fit in any Q1 (of 30) unless you also have Q1Factor higher than 1. So, I *know* that with a Device Buffer Size of 4096 Q1 must be at 14 to be safe (and not ~3 which would be a comparing setting to your Device Buffer Size of 1024).
So, whether you like the sound of a higher Q1 or not, when you set your Q1 to 14 and Q1Factor to 4 and leave all further the same, you won't have any cracks at the end. Now, theoretically your "crack" comprises of four passes (three I think) of wrongly connected buffer data.

Hope this helps. It should ...
Peter

Peter help me understand this in regard to Device Buffer Size(s).  You know at the moment I have to use the FF800 and I can set its Device Buffer Size from 192 - 2048.  Then there's XXHE Device Buffer Size in settings... What are you suggesting I set the FF800 Device Buffer setting at, and then what should I set the XXHE Device Buffer setting?  Also have been setting SFS to 2, Q1 to 7, and Q1F to 2 (which I know is the same as 14 & 1, but I read it here once tried it and haven't changed it)  What do I have to change these settings to?... to get this same awesome SQ

[PeterSt]

Dave, it is not difficult (for me haha). Your FireFace buffer is unrelated to this (well, until you proove otherwise).

So it is just that combination of Q1 x Q1F x DevBufSize which equals 14 x 1 x 4096 (this is not what I said in my previous post, but the result is the same, and I here now used my own settings). And mind you, KS Adaptive Mode ! (can't check Special here).

That you twist the formula so your 2048 physical device buffer of the FireFace is in it, is important for other settings. So, it is good if you sustain that 2048 in there, already not to confuse yourself later. That's how my settings turn into 14 x 2 x 2048 for your situation.

Clear ? but more importantly : does it help / work ?

Peter

[SeVeReD]

Dave, it is not difficult (for me haha). Your FireFace buffer is unrelated to this (well, until you proove otherwise).

So it is just that combination of Q1 x Q1F x DevBufSize which equals 14 x 1 x 4096 (this is not what I said in my previous post, but the result is the same, and I here now used my own settings). And mind you, KS Adaptive Mode ! (can't check Special here).

That you twist the formula so your 2048 physical device buffer of the FireFace is in it, is important for other settings. So, it is good if you sustain that 2048 in there, already not to confuse yourself later. That's how my settings turn into 14 x 2 x 2048 for your situation.

Clear ? but more importantly : does it help / work ?

Peter

sooo, ah
set the FF800 to 2048
&
XXHE setting to 4096

yes?

[SeVeReD]

Dave, it is not difficult (for me haha). Your FireFace buffer is unrelated to this (well, until you proove otherwise).

So it is just that combination of Q1 x Q1F x DevBufSize which equals 14 x 1 x 4096 (this is not what I said in my previous post, but the result is the same, and I here now used my own settings). And mind you, KS Adaptive Mode ! (can't check Special here).

That you twist the formula so your 2048 physical device buffer of the FireFace is in it, is important for other settings. So, it is good if you sustain that 2048 in there, already not to confuse yourself later. That's how my settings turn into 14 x 2 x 2048 for your situation.

Clear ? but more importantly : does it help / work ?

Peter

sooo, ah
set the FF800 to 2048
&
XXHE setting to 4096

yes?

or let me try asking this way.  You have a window that pops up saying to be careful and make the XXHE Device Buffer setting the same as my dac.  So since my dac only goes up to 2048 when I up 44.1 material to 176.4 material, shouldn't I set the XXHE to the same 2048 or should I just set the XXHE Device Buffer to 4096 no matter what the Fireface800 device buffer can go up to?