Jump to content

How to solve sync and latency issues by Introducing Kemper hardware as an outboard gear / external effect in sonar ?


J.War

Recommended Posts

Hello all,

I have some sync issues with reamp tracks using a Kemper Profiling Amplifier (Constant latency ON) and latest Sonar. (Kemper hooked with SPDIF to my RME Fireface UCX lowest latency setting).

There's a latency between the original DI signal and the reamped one, there's a slight latency of 12 tics (at 960 tics per quarter note) or 393 samples. I don't have time to compensate every track by hand.

On the official Kemper forum i was told the following, so before switching definitely to Cubase, i'd like to know if there's a way to do it in Sonar :
 

Quote

Two ways I solve it:

Introduce the Kemper as an outboard gear / external effect if your DAW allows. Here in my Cubase that works like a charm. It has an automatic "Latency Compensation" which detects the latency when you configure the Kemper as an external effect (thus it needs "constant latency" otherwise that would change from profile to profile). With that the alignment is done automatically by Cubase - and by the way super precise. Look out for such an option in your DAW for best quality and convenience.

Manually move the reamped samples and align a well visible transient while zooming in massively. That can come close.

Indeed there will always be a "roundtrip delay" for the way from DAW to Kemper and back. But most DAWs do have tools on board to compensate this as it is absolutely necessary to avoid phasing issues like discussed here.

Constant Latency helps to keep this consistent but does not change the "overall delay" which comes from the latency introduced by the overall roundtrip of the signal out of the DAW into the Kemper and back into the DAW.

 

Best regards

Edited by J.War
Link to comment
Share on other sites

  • J.War changed the title to How to solve sync and latency issues by Introducing Kemper hardware as an outboard gear / external effect in sonar ?

IIRC, Kemper has ~2-4ms of latency (input to output).

There's no way for any DAW software to automatically compensate.

However, if you take the time to configure an External Insert for the Kemper (within Cakewalk/Sonar), Cakewalk/Sonar can measure the latency and compensate for it.  Note this only works if configured as an External Insert.

Otherwise, you have to manually align to the original track.  This is true with any DAW application (not just Cakewalk/Sonar).

  • Like 1
Link to comment
Share on other sites

Thank you very much @Jim Roseberry for the answer.

As you can see on the picture below, i've send my DI track to an AUX with " External Insert ". Inside External insert i've selected my SPDIF out to the kemper as a SEND, then selected the SPDIF in from the kemper as RETURN.

I clicked under " delay " and the latency to compensate shown is 12mS / 594 samples.

I still hear phase issues when layered with the original take and don't know what to do next to make the plugin compensate the desync.

Any help would be welcome.

CAKE.thumb.jpg.fe8ce6916c745f74042f389134d61b7b.jpg

 

Edited by J.War
Link to comment
Share on other sites

First, you need to make sure your audio interface is perfectly aligning audio tracks (when recording).

Most ASIO drivers don't report the actual latency accurately... causing a "record offset".

 

Measure the record offset:

  • Take a really short spike type signal (snare hit, impulse-response, or similar)... and re-record that to a second track via analog output to analog input (physically patch it).
  • Now, zoom way in... and you're most likely going to see the tracks aren't perfectly aligned. 
  • Measure the time difference between the two spikes (measure this in samples.  This is the actual record offset.
  • In Cakewalk/Sonar:  Preferences>Audio>Sync And Caching>Record Latency Adjustment (samples), enter the amount of of the record offset.

Verify the record offset is working:

  • Record the original spike onto a third track (same procedure as above).
  • Zoom way in... and you should see the first and third tracks (spikes) line up perfectly.

 

NOTE:  You should do this for any audio interface (regardless of cost/quality).  You only need do it once.  

Once you've verified all tracks are now properly record aligned, recreate your Kemper External Insert effect (measure the latency).

Kemper re-amps should them line up precisely. 

 

 

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

@Jim Roseberry Thanks for the help.

Using my RME Fireface UCX, i hooked a jack to balanced output 6 and to jack input 3. I then recorded a clip with one sample "spikes" on it. You can see the results below, orange line is the original clip, purple is the newly recorded one :offset.thumb.jpg.bd3f95179325b6e7c5e7638c70ceeceb.jpgne is the newly B

By looking at this, i assume the recorded signal as a latency of 1 sample ? So on the window setting below i should input " - 1 " in the " decalage manuel " box ?

offset2.jpg.5a5b44a837186304ec808b38624765b2.jpg

Is all this still relevant, as i am only using SPDIF ?

Link to comment
Share on other sites

Ok here were are ! i've put " 1 " in the " manual offset " field and made a third take, you can see it below this is the yellow line. (The purple one was the previous one with no offset)

 

offset3.thumb.jpg.1ba18f0f669188805f11ffcc0798d880.jpg

  • Like 1
Link to comment
Share on other sites

@Jim Roseberry after correcting the offset (see above) i did the following :

- KPA is in DI / STACK mode with a dual rectifier profile. So i recorded simultaneously 2 différent tracks : one raw DI  et one with a dual rectifier profile.

- I inserted a SEND into my DI track, i routed that send to a new AUX track.

- In that AUX track i inserted the " EXTERNAL INSERT " plugin.

- On the " External Insert " Plugin inteface, inside " Aux Send " i selected " ASIO fireface SPDIF LEFT ", then inside " Return " i selected " Asio Fireface SPDIF LEFT "

- I clicked under the " DELAY " zone of the External Insert plugin, which gave me the latency to compensate 11mS, 530 Samples.

- I then created a new track and selected the input " ASIO Interface SPDIF LEFT ". i Armed the track and hit record.

- On That new reamped dual recto track, if i hit the PHASE button, it doesn't " sound cancel " the first dual recto track recorded along the DI one and aligning it a bit further by hand make it almost disapear. So i guess the phase is still out (although a lot less obvious thant before) and something is wrong in the process above.

Any idea on what i am doing wrong ?

 

 

Edited by J.War
Link to comment
Share on other sites

If I understand correctly, you're having issues with the reamped (originally DI) signal... and the original Kemper processed signals not aligning.

 

As a quick test, if you do two separate passes reamping the DI signal thru the Kemper, do those two tracks align perfectly (phase cancel)?

 

I would also measure the time offset between the Kemper's DI output and it's processed output.

 

Now that you know that the audio interface itself is recording properly time-aligned, what happens when you do a short test recording (both DI and Kemper processed)... then try reamping that DI signal thru the Kemper?

 

If the Kemper's DI is slightly latent, you just need to figure out the amount.

If that's the case, for the future... use a quality DI box prior to the Kemper.  An all analog path would be zero latency.

If the DI is post A/D, it's definitely going to have some amount of latency.

Link to comment
Share on other sites

Thanks to your suggestions @Jim Roseberry i did ! here is a guitar take in DI / Stack mode.

That guitar take is simultaneously sent UNPROCESSED (DI side) on SPDIF left (upper track) and PROCESSED with the actual loaded profile on SPDIF Right (lower track)

12.thumb.jpg.b9bffb88a6e9070345b313273d317ea5.jpg

By looking carefuly at the transcients, ce can see there's a small latency between the raw D.I side (upper) and the processed side (Lower). On the lower processed track there's a slight delay.

Below is a zoom of one of those transcients, where we can see the processed side had a delay of roughly 8 or 9 samples. Do you agree with me ?

13.jpg.7866d7003fada27545e959ae905bb803.jpg

 

If i understand correctly for a REAMP to be aligned with it's original PROCESSED take, you have to compensate for the reamp loop added latency AND  for the tiny processing latency we just saw above (8-9 samples). Do you still agree with me ?

 

And accordingly to your suggestion, i reamped the exact same way, the same DI two times :

One on TRACK A et one on TRACK B, you can listen here a short sample where i switch the "phase polarity button" of TRACK A several times : https://on.soundcloud.com/Tp9NnSFYAsP63zvx9

The phase cancel is uncomplete ! look below to understand why :

 

Sanstitre-6.thumb.jpg.3111563abf4cd292aa3bfab1d0e2f81f.jpg

This is a zoom of the 2 reamps from the same DI, if i am not blind i think they are perfectly aligned.

But if you look carefuly at those 2 " peaks " at the left of the 2 grey markers, you can see there are slight differences in the waveforms. I think that explains why the phase cancelation is incomplete.

What do you think ?

 

Edited by J.War
Link to comment
Share on other sites

15 hours ago, J.War said:

If i understand correctly for a REAMP to be aligned with it's original PROCESSED take, you have to compensate for the reamp loop added latency AND  for the tiny processing latency we just saw above (8-9 samples). Do you still agree with me ?

Yes, If there's any processing delay from the DI signal, that too would need to be compensated.

 

15 hours ago, J.War said:

This is a zoom of the 2 reamps from the same DI, if i am not blind i think they are perfectly aligned.

But if you look carefuly at those 2 " peaks " at the left of the 2 grey markers, you can see there are slight differences in the waveforms. I think that explains why the phase cancelation is incomplete.

What do you think ?

That's exactly why I wanted you to record two separate passes of essentially the same reamp.

I was thinking they'd likely not be 100% identical (thus won't 100% phase-cancel).

The waveforms are aligned... and they're (obviously) pretty similar.

Link to comment
Share on other sites

You can't expect reamping to phase cancel. Even with digital gear they are emulating analog equipment so there will always be random differences. 

As long as you don't have signals badly out of phase you aren't going to hear negative effects of phase cancellation anyway.

  • Like 1
Link to comment
Share on other sites

@Jim Roseberry Thanks, it's all now cristal clear ! Just one last request,  would you please validate i am using EXTERNAL INSERT plugin the right way to record a latency compensated reamp  :

- KPA in DI / STACK mode. (2 recorded tracks : one from DI unprocessed side and one from processed STACK side.)

- I inserted a SEND into my DI track, i routed that send to a new AUX track.

- In that AUX track i inserted the " EXTERNAL INSERT " plugin.

- On the " External Insert " Plugin inteface, inside " Aux Send " i selected " ASIO fireface SPDIF LEFT ", then inside " Return " i selected " Asio Fireface SPDIF LEFT "

- I clicked under the " DELAY " zone of the External Insert plugin, which gave me the latency to compensate 11mS, 530 Samples.

- I then created a brand new track and selected the input " ASIO Interface SPDIF LEFT ". i Armed the track and hit record.

 

Everything seems done right to you ?

 

@Noel Borthwick Thanks, the official Kemper support confirmed the same info in the meantime.

By the way i can't find a simple way to move a clip by a fixed amount of samples. Like you can do with " Tics " with + or - key in the eventlist.

Let's say i want to move back a clip a amount of 9 samples to realign it with another clip and fix a phase issue, what's the fastest and more simple way to do it ? Channel Tools let's you add a positive amount of samples but not a negative amount.

Edited by J.War
Link to comment
Share on other sites

5 hours ago, J.War said:

Channel Tools let's you add a positive amount of samples but not a negative amount.

Did you try using the - in the number row on your keyboard? IIRC, the - on the NumPad is hard coded into the program and 1) highlights a cells contents and 2) decrements any existing value by a single whole number (with each press). You have to watch the cell contents to see this behavior, but entering "-3" with the NumPad will often come up as "3." For years I thought all of those affected cells wouldn't accept negative numbers.

Edited by mettelus
Link to comment
Share on other sites

Thanks for the hint, unfortunately entering "-3" would come up as "0".  Even tho, i've found no way to make to value goes backward either with the numpad, or either by moving knobs...

I've found Waves InPhase can do it by tweaking the " SM " cell just under the " MS " cell. It then shows the positive or negative samples amount. 

Link to comment
Share on other sites

Did you try using the number row on your keyboard? It is possible that field does not accept negative numbers (not sure), but if it will, the - in the number row (not the NumPad) should work. If entering -3 comes up as 0 (totally different number), it is quite possible that field will not accept negative numbers 😞

Link to comment
Share on other sites

7 hours ago, J.War said:

 

@Noel Borthwick Thanks, the official Kemper support confirmed the same info in the meantime.

By the way i can't find a simple way to move a clip by a fixed amount of samples. Like you can do with " Tics " with + or - key in the eventlist.

Let's say i want to move back a clip a amount of 9 samples to realign it with another clip and fix a phase issue, what's the fastest and more simple way to do it ? Channel Tools let's you add a positive amount of samples but not a negative amount.

You can try using snap by and setting the snap amount to samples and enter the sample value. Then zoom in and move the clip left till it snaps to the amount you need. Nudge is probably an easier way to do it.

image.png

 

  • Like 1
Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...