TOPIC:

dreite wrote:

You probably do have the necessary electrical alternative.  (You just don't realize it.)
A basic two-channel soundcard, 
 

My laptop doesn't have any analogue audio inputs to feed into whatever performs soundcard functionality on it.

What I proposed really is 100% fit for purpose for what is relevant here. You questioned whether the information MSB supplied was accurate, presumably as it was such a surprisingly large number. I was suggesting confirming it because if it is accurate then the OP needs a serious rethink of their plans. Being certain of the facts therefore seemed prudent.

I did question it.......until he mentioned he talked directly to a guy at MSB.
I think that renders the whole issue moot, at least from his point of view.

Well, I think you need to invest a couple of bucks and outfit yourself with a proper measurement setup.
A good, two-channel setup (maybe a Focusrite 2i2) shouldn't set you back more than $150.  And much cheaper ones would work fine as well.....at least for this type of thing since we're not concerned with noise floor or distortion or etc.
Many years ago I used SpectraPlus for a software setup, but now my preferred go-to is ARTA.

Dave.

@dreite - I'll bear your recommendation in mind but currently I really would have no use for an external soundcard to justify the cost. If I were to spend extra money on new measurement gear right now it would actually be a UMIK-2 that would most interest me.

@rortmanns - it's up to you whether you wish to confirm for yourself what is actually going on or not. If the latency of your MSB DAC really is such that you'll find your subwoofer signal arrives hundreds of milliseconds before that from your main speakers then this simply won't sound good. You'd then have four main options, of which I expect the 4th will appeal most:

1. Don't use use MSB DAC.
2. Use your MSB DAC but only in its low latency mode intended for TV etc.
3. Don't use your subwoofers.
4. Find an additional piece of audio equipment to place between the SHD and your subs to apply the required compensatory delay. I don't know of one but I'm sure such things must exist, most likely made for the pro audio market.

For the record I don't discount that your MSB DAC sounds better than your SHD's own DAC by the way so I wouldn't just push the first option as the simple and optimal solution..

Ultrasonic, this is getting quite tiresome and pedantic now.  

I think @rortmanns is well aware of the problem with using a 500ms relative delay in one branch of his system.
He can sort this out however he sees fit.

Dave.

dreite wrote:

Ultrasonic, this is getting quite tiresome and pedantic now.  

I genuinely have no clue what you're referring to. I just posted for @rortmanns benefit, which I'd suggest you focus on too.

Ultrasonic wrote:

rortmanns wrote:

Is the latency of the internal DAC to the SHD known?  The MSB delay is about 0.5Sec, MSB advised me that’s about the figure.

In answer to the first question - no, I don't believe it is. There are figures for the combined latency of the whole input-output signal path (the 25 ms mentioned above) but the vast majority of that will be from the DSP section that would apply to the signal sent to your MSB DAC just the same as via the SHD's DAC and onto your sub. The latency of just the SHD's DAC will be much less than this total and won't pose any issue at all though, and would just get included in any acoustic measurements.

What would be a HUGE problem though is if your MSB really does have a 0.5 s (500 ms) latency. That would basically make it impossible to properly integrate your subwoofers, since the maximum conpensatory delay you could apply to your subs on your SHD is 30 ms, and I'd be stunned if your subs gave you the capability of setting an additional 470 ms delay. Relative positions of main speakers and sub in your room will affect time delay requirements but would be utterly dwarfed by the 500 ms of your DAC.

Your number one priority should be to make a measurement to confirm that the latency difference between the DACs really is as much as ballpark 500 ms, because if it is you need to totally rethink what you're doing. To do this I would use REW to make the following measurement:

• Set your microphone up at a point equidistant from your two speakers*.
• Set up cabling and signal routing in your SHD so that the SHD DAC is used for, say, the left speaker and your MSB DAC for the right speaker.
• Do not include your subwoofers for this test.
• In the measurement window, select to use an acoustic timing reference and set this to be your left speaker (the one using the SHD DAC). 
• Make a measurements for your left speaker on its own, and for your right speaker on its own (sweep up to 20 kHz).
• Once you have the measurement, choose the overlays option, and then impulse response. The signal for each speaker will have a sharp initial spike (the first arriving direct signal from the speaker) followed by a sequence of smaller spikes (which are reflected signals from room boundaries and things in your room). You'll need to zoom out A LOT to see the signal for your right speaker if there really is a 500 ms delay. The difference in latency between the MSB and SHD dacs will be the time delay between the initial spike from your left speaker (will be at 0 ms on the graph) to the initial signal from your right speaker (which will be at about 500 ms if your info. is correct, but hopefully it's rather sooner!)

*To check that the delay truly is ballpark 500 ms just setting up with a tape-measure will be fine, and you main listening position should be equidistant so I'd just use there. For a more accurate measurement of latency above you could check this acoustically using the SHD DAC for both channels first, and even compensating for the difference on the main measurement. What matters here is an order of magnitude check though. 500 ms is unusable, whereas say 5 ms would be.

I’ll give this a try.  I’m pretty confident that MSB gave me the correct advice because I have since found another thread with a user trying to do a similar thing with an MSB DAC.  I’ll still do the test because I’m interested to know what the ‘video’ mode (supposedly low latency) delay time is and just to get an accurate number rather than the ‘about 0.5 sec’ I was advised.  The user in the other thread found a MiniDSP unit that was 8 channel and had up to 3 seconds of delay.  That’s an option for me or as stated before, I can redeploy this DAC in my home theatre room for stereo playback, the Wilson Audio speakers don’t really need subs.

I need to let the SHD DAC burn in properly, see what it really sounds like (assuming it improved as many components seems to do).

Im a bit slow trying things, I recently had major surgery, I can’t lift over 5kg and I get a bit tired.

Ultrasonic wrote:

@dreite - I'll bear your recommendation in mind but currently I really would have no use for an external soundcard to justify the cost. If I were to spend extra money on new measurement gear right now it would actually be a UMIK-2 that would most interest me.

@rortmanns - it's up to you whether you wish to confirm for yourself what is actually going on or not. If the latency of your MSB DAC really is such that you'll find your subwoofer signal arrives hundreds of milliseconds before that from your main speakers then this simply won't sound good. You'd then have four main options, of which I expect the 4th will appeal most:

1. Don't use use MSB DAC.
2. Use your MSB DAC but only in its low latency mode intended for TV etc.
3. Don't use your subwoofers.
4. Find an additional piece of audio equipment to place between the SHD and your subs to apply the required compensatory delay. I don't know of one but I'm sure such things must exist, most likely made for the pro audio market.

For the record I don't discount that your MSB DAC sounds better than your SHD's own DAC by the way so I wouldn't just push the first option as the simple and optimal solution..

Should have read all the replies before responding earlier.

You’ve summarised my options nicely.  The MSB is a fabulous sounding DAC but I’m a pragmatist, I may just replace it, that’s the other option.  The issue then becomes one of not knowing the latency of the replacement.  The SHD only gives you a 30mS window which isn’t much, that’s not its purpose.  At the end of the day if the SHD doesn’t meet my expectations it will be replaced.

As for measuring the timing, I’m happy to use the speaker measurement method to do this because I need to setup the Umik-2, REW and the USB connection to the SHD for trying the MSO process and trying out Dirac in any case.  To change over the SHD DAC from the MSB is just moving 2 XLR connections, it’s real easy.

The electrical measurement process is a bit more difficult.  I’d need to stumble through my pig sty of a shed and dig out my oscilloscope and I’d still need to set up everything else anyway.

Over Xmas I’ll do the measurements and I really want to hear the speakers with the subs optimised and then Dirac.  If I like the outcome I’ll simply redeploy the MSB and get on with treating the room slightly.  I really just want to get back to the music.  I can’t do much in the other room with the MSB, I’m waiting on the mains power amp, I’m hoping there’s a little guy in Japan soldering it up at the moment, there still seems to be a big delay with some audio components.

Id like to thank everyone for all the help.  I was hoping this was going to be much easier than it has been, if only manufacturers published better spec data.

rortmanns wrote:

The electrical measurement process is a bit more difficult.  I’d need to stumble through my pig sty of a shed and dig out my oscilloscope and I’d still need to set up everything else anyway.

Apparently not though since @ dreite  has mentioned 3-4 times how easy his method is (presumably not needing an oscilloscope either). Dave, rather than keep saying your electrical method is easy, could you not just describe it? This thread would seem as good a place as any to do so I'm sure a few onlookers (like me) would be interested to learn. 

Ultrasonic wrote:

• Set your microphone up at a point equidistant from your two speakers*.
• Set up cabling and signal routing in your SHD so that the SHD DAC is used for, say, the left speaker and your MSB DAC for the right speaker.

Hi @Ultrasonic, I read interesting posts like this as it helps with picking things up. I see Dave isnt a big fan though
I understand what you're saying about this giving the OP an understanding the size of latency he's getting but accuracy will rely on how equidistant the mic set-up is from speakers. Could you take a second measurement to help? basically reverse the speakers so SHD on round 2 = Right and MSB = Left. Then average both SHD measurements and both MSB measurments then look at the difference? Would that remove the innacuracy of the mic positioning?

 duplicate

1.25 feet is equal to 1mS if I recall correctly.  Any distance error given the large expected latency makes is probably irrelevant.  I use a digital distance measuring thingy to set my speaker distance from the walls and listening position. The measuring thingy is a Disto D510, makes life easy.

@asx77  Yep, I'll make a short video showing my testing scheme and post on my website when I get a chance.
It's very straightforward.

Dave.

rortmanns wrote:

1.25 feet is equal to 1mS if I recall correctly.  Any distance error given the large expected latency makes is probably irrelevant.  I use a digital distance measuring thingy to set my speaker distance from the walls and listening position. The measuring thingy is a Disto D510, makes life easy.

yes - fair enough, you'd need to be a daft amount off equidistant!

rortmanns wrote:

As for measuring the timing, I’m happy to use the speaker measurement method to do this because I need to setup the Umik-2, REW and the USB connection to the SHD for trying the MSO process and trying out Dirac in any case.  To change over the SHD DAC from the MSB is just moving 2 XLR connections, it’s real easy.

Yes, what I suggested being an extension of the sort of thing you'd be doing anyway was part of my thinking. I also got the feeling from your posts that you'd be curious to see what is actually going on like I would, and like you I'd personally compare to the DAC's low latency mode. 

asx77 wrote:

Hi @Ultrasonic, I read interesting posts like this as it helps with picking things up. I see Dave isnt a big fan though
I understand what you're saying about this giving the OP an understanding the size of latency he's getting but accuracy will rely on how equidistant the mic set-up is from speakers. Could you take a second measurement to help? basically reverse the speakers so SHD on round 2 = Right and MSB = Left. Then average both SHD measurements and both MSB measurments then look at the difference? Would that remove the innacuracy of the mic positioning?

I did refer to this above but what I'd do if I wanted to make the measurement as accurate as possible is to first do a measurement at the MLP using the SHD DAC for both left and right signals. Any amount the right channel arrives before or after the left (the timing reference) I'd take to be due to a geometric setup difference which would be the same when the MSB DAC is swapped in for the right speaker. So let's say the right speaker signal arrived 0.5 ms later on the first test I would then subtract 0.5 ms from the difference measured with the MSB DAC in use. But as above, a setup error would have to be pretty large to somehow mask what we're looking at here!

I just checked back out of curiousity and for last acoustic measurement I made at the MLP in my setup the arrival time difference between left and right speakers was 0.004 ms (4 microseconds), and the time before that it was 0.020 ms. In both cases this was measuring to place my microphone once, not repeatedly moving it around to optimise its position to minimise the arrival time difference. Just making the point that it's perfectly possible for these differences to be totally negligible, not that it's necessary.

There’s another option I’m going to try.  I could just run the mains and the subs from the DAC direct and use the crossover built into the D15S subs, the sub has delay (100mS) and 9 PEQ plus phase control.  So I could run the mains full range and adjust the subs with MSO settings for PEQ etc.  I’d only be using the SHD for Roon and Dirac then.

So another option.  I’m going to try that by just turning off the high pass on the mains for now, saves me making more XLR cables.