Switched from Logic Pro to cubase. I created a template for analog summing. Tracks can be assigned to any of 4 stereo groups.
In testing I have something odd… I created a mono click track - output is routed to input of 2nd track. Recorded track is early. Any ideas?
Misreported latency can cause this. You can check the difference in samples and put it in the record shift box…check a few times to make sure it’s a stable amount but note that it will likely change at different sample rates.
Thanks, how can I check the difference? Logic Pro has a latency measuring plugin. Does Cubase have a tool to measure latency?
You can set up an “External FX” loop out and back and ping it to measure or just change the ruler to samples and zoom in to read it off manually.
Are you trying to test input latency through your interface using the internal click (and not using latency compensation on the track)? Ok, I think there’s a better way to test your interface’s ability to
align your input to the click. Here, try this:
Generate the ‘audio’ click via the Cubase option under the '‘Project’ tab in your project. Open it and scroll down to the heading ‘Signature Track’. Click on this (after you’ve created your Project Locators), and then click on ‘Render audio clicks between locators’. Having done this, you can be assured that the audio click track is mated perfectly to your time signature. Now you can try to record either a MIDI instrument or instrument track to see how close your timing is to the click.
I did something similar - I created a drum track with a 1/8 note hi-hat pattern, then rendered it to a new track. Then I created a new mono track and routed hi-hat mono out patched to input of the new channel… Record… New track looks like 5-6 samples early. I don’t know how to precisely measure the offset.
Thanks - How do I ping?
Ok, I created an external FX bus - when I pinged it, it measured 5ms… That seemed like too much latency for my system - Lynx Aroura (n) Thunderbolt. I then tried adjusting the Record Shift in the Studio Setup with -1 samples… That looks very close. I wish there was a way to measure the precise latency. At 0 samples Record Shift it is recorded early, which is strange.
TBH, the DAW is not a time machine, so it could not record early.
That means something delays something too much.
But your 2nd screenshot showed a latency of 5ms.
What is your exact routing?
Do you use export or recording for the external FX?
Can you provide screenshots from the mixer view with the routing shown?
The ping should be the precise latency overall (but I’m not sure if it’s 100% sample accurate tbh). But bear in mind the ping bypasses the usual delay compensation so the amount you want in record shift to be used in general is actually the difference between the ping and your reported latency which looks to be around .4ms…maybe in the region of 15 samples.
I would just run the loopback test with a sharp percussive sample and read the difference off on the ruler set to samples (zoomed in as far as needed)…this is what I’ve always done.
1 Like
TBH, the DAW is not a time machine, so it could not record early
It doesn’t record early. It places the parts early.
1 Like
Thanks Grim, I was able to get recorded audio to align by adjusting the Record Shift to -1 samples. I’ll test to see if that changes with multi-track recording. I opened a support ticket with Steinberg to find out how Cubase determines inut and output latency in the Studio Setup settings… There is a record shift happening even with all setting at zero.