Topic: Totalmix "Centre" Phase cancellation

I have a HDSPe MADI FX with Windows 7 Professional. I am not new to TotalMix FX, but I   have recently disconnected my AES monitor controller/DAC and MADI devices and I am now only using the Integrated TRS Line Out/Headphone output.

I have updated the drivers and flashed the hardware and I have re-read the user manual and made some adjustments to the DSP settings accordingly - I made the WDM Phones channel the only "Speaker" and set it as the default device in Windows/Sound.

Interestingly I resolved the apparent routimg issue at one point, but when I disconnected my headphones and connected  another pair of stereo speakers via a TRS to 2x XLR cable, the problem returned and plugging the same headphones back in didn't reverse things back to function properly.

What I am seeing manifest is the stereoscopic information (the sum of the difference between L and R) is completely phase cancelled. I can only conclude that there are duplicates of the same signal that mustn't be panned hard L or R. This is supported by the fact that if I phase flip one side at a time on the main output, the centre information restores and only half of the duplicate signal reaches the main fader.

I've tried all kinds of routing configurations and reset all submixes various times etc.

I can't see activity on any channel other than the Phones Software Playback, which I have included in the Main Output submix.

I must be misunderstanding something fundamental here... there are no signals entering any hardware inputs.

Please help! Cheers

Re: Totalmix "Centre" Phase cancellation

With what software and what audio are you playing/working. Does it also happen with a simple mp3 playback from windows media player?

Vincent, Amsterdam
https://soundcloud.com/thesecretworld
HDSP9652+ADI-8AE  2X HDSP9632
Cubase pro10

3 (edited by Polym0rphed 2020-01-09 17:15:19)

Re: Totalmix "Centre" Phase cancellation

vinark wrote:

With what software and what audio are you playing/working. Does it also happen with a simple mp3 playback from windows media player?

Hi Vinark - I first limited my testing to mp3 files in winamp or vlc, yep, but I am also a long time user of Cubase 7.5 (which I haven't updated) and if I add the stereo outputs and allocate them to PhL and PhR, I should expect any sub bus that is routed to the main fader to recieve those signals and they do, but the same phenomenon exists where panning anywhere in cubase hard left or hard right actually phase aligns the centre image.

Re: Totalmix "Centre" Phase cancellation

try totalmix reset and/or have a look in the matrix mixer of totalmix. Could be a loopback activated maybe.

Vincent, Amsterdam
https://soundcloud.com/thesecretworld
HDSP9652+ADI-8AE  2X HDSP9632
Cubase pro10

Re: Totalmix "Centre" Phase cancellation

Hello!

It sounds like the Mid Channel of a Mid/Side configuration is muted, to me. So can you check if you have M/S Processing activated in any of the Software Playbacks Outputs?

Here's a quote from page 71 of the manual:

27.7 MS Processing The  mid/side  principle  is  a  special  positioning  technique  for  microphones, which results in a mid signal on one channel and a  side  signal  on  the  other  channel.  This  information  can  be  transformed back into a stereo signal quite easily. The process sends  the  monaural  mid  channel  to  left  and  right,  the  side  channel too, but phase inverted (180°) to the right channel. For a better understanding: the mid channel represents the function L+R, while the side channel represents L-R. During record the monitoring needs to be done in 'conventional' stereo. Therefore TotalMix also offers  the  functionality  of  a  M/S-decoder.  Activation  is  done  in  the  Settings  panel  of  the  Hard-ware Input and Software Playback channels via the MS Proc button. The  M/S-Processing  automatically  operates  as  M/S  encoder  or  decoder,  depending  on  the  source  signal  format.  When  processing  a  usual  stereo  signal,  all  monaural  information  will  be  shifted into the left channel, all stereo information into the right channel. Thus the stereo signal is M/S encoded. This yields some interesting insights into the mono/stereo contents of modern music productions. Additionally some very interesting methods of manipulating the stereo base and generating stereo effects come up, as it is then very easy to process the side channel with Low Cut, Expander, Compressor or Delay. The most basic application is the manipulation of the stereo width: a change of the level of the side channel allows to manipulate the stereo width from mono to stereo up to extended

I don't have a MADI FX, but this is what I can figure from your post.
Hope it helps!

6

Re: Totalmix "Centre" Phase cancellation

Nice. Another theory: ground connection is missing - broken TRS socket.

Regards
Matthias Carstens
RME

7 (edited by Polym0rphed 2020-01-10 09:49:36)

Re: Totalmix "Centre" Phase cancellation

Thank you all for the replies - yes it is presenting like the mid channel in M/S processing is muted, however, I looked at all the relevant channels with this in mind, which helped me understand that the problem lies elsewhere despite sharing the same symptoms.

The matrix view looks exactly as I would expect it to after a submix reset and I couldn't see any routing I wasn't expecting. What would a loopback look like in the Matrix, just in case?

At this point, the only theory I haven't tested is a possible problem with the socket itself (I've ruled out the male jacks by testing those cables/devices in other systems.

The socket does have a little bit of play, but I wouldnt say it's "loose". I don't have any instruments to test the signal current physically though hmm

I guess I can take the card out for a closer look at least. Any suggestions?

Re: Totalmix "Centre" Phase cancellation

If with headphones and a little wiggling you can get it to work correctly, if even for a moment, you know it is the socket.

Vincent, Amsterdam
https://soundcloud.com/thesecretworld
HDSP9652+ADI-8AE  2X HDSP9632
Cubase pro10

Re: Totalmix "Centre" Phase cancellation

I dug up my physical user manual for the HDSPe and read through the whole thing again (I find it easier to maintain concentration when reading from paper). There are a couple of features that I admit I've never used or realised existed, but nothing that should cause what I've described.

I will double check everything, including wriggling the jack and report back as I'm sure the thread will help others in future.

By the way, I read that the Phones output is not inherently linked to any physical input thus rendering it incapable of loopback.

Another couple of observations - I connected an analogue stereo meter bridge to have visual inrormation to refer to after the main output and I noted that no matter how I route or pan things (and even when the meters in TMFX show activity ONLY on one channel), the meter shows the same signal on both channels simultaneously. - I'm not sure if this helps narrow down the possible causes, but in my mind it is pointing towards a hardware defect (as suggested, within the socket).

10 (edited by Polym0rphed 2020-01-21 15:21:43)

Re: Totalmix "Centre" Phase cancellation

I re-tested the headphones that gave me a full L or R (opposite silent) signal once before and realised that the plug has a loose connection; the true signal being output appears stereo on both L and R - in other words hard panning (or solo/muting) doesn't result in total silence in the opposite side.

The routing is correct in matrix view and there is no signal activity on any channel other than the headphone output in question and I've tried submixes where the Main out is unassigned, assigned to PH or assigned to either MADI or AES - every option gives the same results.

I've tred various configurations in Windows sound settings, such as setting other audio devices to default, setting AES, MADI 1/2 etc, PH to default/or left unassociated.

All combinations yield the same results... I manually checked all in/out channels including mains for enable/disable status for mid side processing and all were/are off. Phase flipping  doesn't cause total cancellation when done on hard panned mono pairs. It's like there are two layers of signal or that the left and right channels are each subdivided into a further stereo pair that is entangled (can't be mono split because it is summed back to the single channel panner in TM) - I don't get it. The connection of the poles in the socket seems consistent and fine.

* In the cases where I had other audio devices set as default, I either used Cubase 7.5 or various music player softwares that allow manual allocation of IO to test or confifure audio

Re: Totalmix "Centre" Phase cancellation

Hey Polym0rphed!
Did you figure out?

As MC has, allready, suggested, it must be ground connection missing!

So, have you tried with different pair of headphones?
Is the card firmly seated in the PCIe slot? (could loose ground, if a few pins are not seated all the way in)
Is your motherboard screwed in with every screw required?(Screws in motherboards provide grounding)
Is your PC's PSU and power cable in good condition?

These are a few things that comes to my mind, you could check, if you haven't allready!

Re: Totalmix "Centre" Phase cancellation

MetalHeadKeys wrote:

Hey Polym0rphed!
Did you figure out?

As MC has, allready, suggested, it must be ground connection missing!

So, have you tried with different pair of headphones?
Is the card firmly seated in the PCIe slot? (could loose ground, if a few pins are not seated all the way in)
Is your motherboard screwed in with every screw required?(Screws in motherboards provide grounding)
Is your PC's PSU and power cable in good condition?

These are a few things that comes to my mind, you could check, if you haven't allready!


Agreed - it is a grounding problem. I'm not at the location where all my gear is so I haven't had a chance to try various headphones out - the two sets I have on me have loose connections themselves.

I'll crack open the PC and inpsect the card coupling etc. All the other leads are new and the PC was constructed by a Pro Audio business, so I can't imagine any defects in its assembly, but I'll look around.

The phenomenon is basically exactly what would happen if you used a stereo to stereo Y splitter cable - to sum from line level TRS stereo to balanced L+R mono, the cable should be wired with resistors - the circuit is different.

The cable I have is supposed to have such a circuit- I suppose the resistors are within the sealed Y junction part way down the cable (at each connector there is no evidence of resistors, not even the high ohm bridging "fuse" resistor you'd expect to see at the mono plug end. OR it is actually a regularly wired splitter...

13

Re: Totalmix "Centre" Phase cancellation

The cable that you have (according to your first post) is a simple TRS (stereo) to 2 x XLR Male (Mono L/R), and will not have any resistors as they have no purpose here. It might be possible that the cable is doubling the TRS into 2 x XLR by connecting Tip and Ring to pin 2 and 3 respectively. That would be a very unusual, and for this application wrong cable.

Regards
Matthias Carstens
RME

Re: Totalmix "Centre" Phase cancellation

MC wrote:

The cable that you have (according to your first post) is a simple TRS (stereo) to 2 x XLR Male (Mono L/R), and will not have any resistors as they have no purpose here. It might be possible that the cable is doubling the TRS into 2 x XLR by connecting Tip and Ring to pin 2 and 3 respectively. That would be a very unusual, and for this application wrong cable.

I thought I had sent my reply but lost it.

See, when I purchased the cable I specifically asked if it was just a Y cable or if it had resistors in the circuit etc. He assured me it was a passive TRS to L+R cable. Clearly it isn't.

I could strip back this cable and add in the necessary resistors OR make a junction box with appropriate I/O...

The output is 75ohms +13db via TRS 6.3mm.

The rane circuit link I've used in the past is broken sad Do you know what resistor values I'd need and for which pins? Cheers!

15

Re: Totalmix "Centre" Phase cancellation

We all lost overview meanwhile, this thread has become somehow useless. And now you even misread my last reply. You do NOT need resistors in a cable, forget that. And I am not aware of any such cable that has them inside. Also already in your first post you wrote that your phones show the same issue. So how can it be the cable? It seems the TRS socket on your card is broken. Get yourself a technician with a multimeter, he can quickly clear a missing ground issue - and check your cables as well. Just takes a minute...

Regards
Matthias Carstens
RME