Set a loop over a selection
Set the punch points as a subset within that loop range
Option to store takes in a single track
(My test was using four tracks)
Record multiple times through.
Expected: the audio from all takes lines up
Actual: each subsequent take is a little later than the previous
 
Could there be a setting I have messed up?
No input monitoring, no effects.
 
Steven

Are you  looping audio or midi tracks?  I've had a problem looping midi tracks (as that's primarily all I'm using while looping) for every version of Sonar I've used dating back to the original, I believe.  So incredibly annoying to set up a kick & snare in a clip, for example, and some hi-hats on another, and listen to them drift further apart with each successive loop.  Doesn't matter if the clips are regular or converted to step sequences, the drift always happens.

There's nothing in the settings I'm aware of to remedy this problem, but like you, I wonder still if it's something I've overlooked (thought I doubt it). 

I've recently toyed with the idea of making Sonar my primary sequencer but things like this it frustrating.

I'll start a thread of my own if we don't get any satisfactory answers in this one.

This was just audio, no MIDI. 

When I was checking the X1 fix for ASIO loop recording without punch-in recording, I also checked using loop with punch-in and found some weird results. So I wonder if older versions of punch-in worked, and the latest fix for loop recording broke punch-in. http://forum.cakewalk.com/tm.aspx?m=2508189

I did some measuring and some math.  Found that the offsets are about 350 samples which is the amount SONAR indicates as the offset for my sound card.  Hmmm...

http://forum.cakewalk.com/tm.aspx?m=2564955 solution was to use groove clips.

Cactus Music


http://forum.cakewalk.com/tm.aspx?m=2564955 solution was to use groove clips.

Not true for me.  I've tried regular clips, groove clips and  step sequences.  I double-checked the groove clips to make sure they're the same length.  They still drift noticeably about the 5th or 6th time repeated, getting progressively worse.  I have an i7 - processing  power is not the issue.  Maybe I'll mess around with some of the buffers.

Just to be clear.  I am simply recording audio.  The transport is looping.  I am not creating loops.  In this instance I was creating a "choir" effect by singing the various parts by standing in different places in the room while the playback looped over the chorus.

Afterwards when I looked at the lanes on each track I saw that each take was a bit later than the one before it. 

So I recorded the click track and did the same thing (to rule out the human factor).  That's how I was able to measure the offset.

Steven Bell, since you've developed a procedure to quantify the error, we users would be grateful if you would submit a bug report! A workaround for you might be use loop recording without using punch-in. After you have your multiple takes in layers, you can bounce them to your chorus and then trim the two ends.

bvideo


A workaround for you might be use loop recording without using punch-in. 

In my experience, any loop recording (with or without punch-in) causes this problem.  I posted a similar question a few days ago and the suggestion was made to change the record latency offset to better match the latency of the interface. Although the first track does line-up better,  the problem still existed on later looped tracks.