I would love this, since all my inputs are not the exact same. Some are directly from my interface and need 0 samples of shift, but others come from a spdif or ADAT input that have to go through an A/D conversion stage that introduces 60 samples (1.3ms) of latency. Right now I have to constantly go into my settings and change the Record Shift depending on the input I use to record. I also cannot use both types of inputs at the same time. It would be awesome if in the F4 Audio Connections menu you could add a custom Record Shift to all the added inputs.
Hello and welcome.
Cubase usually allows to only use one audio device and it is that device’s driver that must synchronize all incoming streams.
Which audio device are you using? By any chance an aggregate device on Mac?
I use a UAD Apollo Twin, the native preamps of the interface necessitate no Record Shift. On the same interface I use the spdif input with a Behringer Ultramatch to convert two outboard preamps. I also use the same digital input of my interface with an ADAT signal from a Scarlett OctoPre to input 8 preamps for larger sessions. They all appear in Cubase under the same interface with one single driver, just different inputs. The problem is that the Behringer and the Scarlett have some latency in their A/D conversions (e.g. around 60 samples for the Behringer). I sometimes use these inputs at the same time as the two Unison analog preamps of the Apollo. So the general Record Shift is not enough to adjust the latency of all the devices. I would need to be able to adjust the Record Shift per input.
I am not aware of any settings in Cubase that allow for a delay per Input Bus or Device Port.
There is track delay, of course. You’d have to set that parameter on every track that you record with the Behringer, and it is in milliseconds.
I could also imagine that using shorter audio cables (mic, guitar) for the Behringer than for the Apollo might help.
Maybe somebody else has a better idea than me.
There is no setting like that in Cubase, exactly, that’s why my title says “Requested Feature” 
I figured that maybe a dev would see it and it might make it as a feature in a future version of Cubase.
For now I do what you suggested, I set a negative track delay of 60 samples (-1.3ms) when I use the Behringer for example. Or I just set the Record Shift to 60 samples when I use only those preamps, etc. It’s not an optimal workflow. The track delay means the audio clips are not aligned to their “real” position, and changing the Record Shift is a lot of button presses if I’m moving around a lot between tracks.
Welcome, @David_Boily !
For this, it’s then better to use the Feature Request tag (after a [click] on the pencil, on the right of your thread title) :
1 Like
Have you contacted UAD about this?
I’ve used similar setups (expanded the analog i/o via ADAT) and did not experience any timing discrepancies between the native inputs and the ADAT connected ones. Different manufacturers in my case however.
The latency you’re dealing with depends on the outboard gear you plug in. Most of the time, you don’t really notice it, especially if you’re just singing or playing a guitar part, because a latency of 1.36ms isn’t noticeable. However, with a loopback from another track, the latency becomes more obvious. For example, if I send a track to the LINE4 output from the Apollo, into my ISA One DI, through my Stam Pultec, into my Behringer Ultramatch, and back into the Apollo via SPDIF. You can then easily measure the latency of the signal path.
The ISA One and Stam Pultec are fully analog, so they don’t add any latency. The Behringer, however, adds latency during the A/D conversion, and the Apollo might also contribute by processing the SPDIF signal. This specific chain results in 60 samples of latency. Previously, when I used a Presonus A/D converter, it gave me 174 samples of latency. This is based on a 44.1kHz sample rate, and I suspect different rates would yield different results.
The latency isn’t an issue with the Apollo or Cubase itself. The delays originate in the outboard gear, which Cubase or the audio drivers can’t be aware of. I assume Cubase’s “Record Shift” option exists to compensate for these kinds of latencies, since it needs to adjust by -60 samples, something that can only be done after recording, at the track or audio clip stage.
This setup works fine if I only use outboard gear or integrated preamps, but my current workflow requires me to either adjust the Record Shift depending on the input I’m using or manually nudge the audio clips if I use both types of inputs simultaneously. I thought if there was an extra Record Shift option per input, it could automate the process. It doesn’t seem like a complex feature to implement, e.g.
ClipStart -= RecordShiftByInputMap[inputX] + GlobalRecordShift;
I would keep the global shift for convenience and just add an extra option in the input mapping under the F4 Audio Connections panel, which has ample space in its widget to accommodate this new setting.
I haven’t really given this any thought so far. But I took your post as a reason to test my own gear. Seems like I have a discrepancy of some 52 samples between converter devices.
Good to know.
Thanks for bringing it up.
1 Like
It depends. If you e.g. are moving the same source with two mics, that latency will result in phase issues and become quite noticeable.
But then you’re measuring round trip which is not exactly the same.
If you haven’t already, I would try feeding one input on each of you A/D’s with the same (external) signal. I have done this. My test was not very scientific as I used two identical microphones side by side rather than perhaps a signal generator. But a discrepancy of 1+ms would have shown up.
1 Like
That’s great point! I had not tested that. 
I just tested it with some hand claps through an SM57 fed to the Apollo internal preamp and the ISA One (bypassing the Pultec but going through the Behringer AD converter) at the same time with a signal splitter. I can confirm that the signal path going through the Behringer (even with a different preamp, with or without the Pultec, etc) has 60 samples of latency compared to the internal Apollo preamps.
My goal in testing the loopback was to make sure that my performances when I do overdubs (guitar, drums, singing, etc) are not desynched compared to what I heard while recording. I also get 60 samples when I just record the sound of the headphones back into Cubase. I mean 60 samples is just 1.36ms, so in the sense of my overdubs not being in time with the other tracks it’s not a huge deal, but it’s not nothing.
But, exactly as you said, in the sense of a multi-mic situation (stereo guitar, piano, Glyn John 3-mic drums, etc) then 1.36ms is enough to ruin things.
I cannot be the only one that could use this feature. There are some preamps that have their own A/D converters (e.g. the ISA One Digital). If you have more than one type of converter in your multiple signal chains, there a good chance their latency is not the exact same.