I won’t certainly position myself as a genious of music assisted by computer but, may be are you so used to your way of working with Cubase and the nature of your produced music that you can’t understand ?
No, this is the trail effect that sometime we like to have on guitar sound, so no fragmentation of file in this take, but tempo need to change. It could certainly be re-constructed using, as you say, cutting and pasting, mixing, etc. but not the real thing, just result of a digital tool as Microsoft Excell (I appreciate) used to compute a salary.
Simple constat for the benefits of storing tempo (not the tempo track, just tempo changes) in each physical file :
no more average value, source of errors when you use time tools on the file.
when the file is moved between projects, it keeps its tempo changes and could be adapted easily to new context.
all the existing tool will continue to work but they have to consider subsections of each file.
And concerning the statistical point of view in these forums concerning new features, I hope that Steinberg evolution path is not build only from these forums, nice but sometime limited by helpful guardians of the faith.
But then what would happen if the tempo change is a ramp that increase/decrease over time, and not just jump points ? So yes, it should basically store a snapshot of the tempo track…
But Cubase already has plenty of features to achieve that, like tempo track import, tempo detection, time warp, etc.
I don’t know of any sample libraries that feature samples other than fixed tempo and fixed time signature. This is for an obvious reason, as it has to be coherent and compatible between every DAW.
I can not disagree with you as you have some good points, but then that would imply to make changes to the WAV format so that it can store specific information that can only be used by Cubase.
Since Steinberg doesn’t own the WAV format, and as WAV is a standardized format that has been designed to work seamlessly between all operating systems and programs, I don’t think this would ever become possible.
It’s digital science, simply storing each point in the ramp would be ok, but it could be simplified/compressed certainly just as mpeg.
tempo detection is impossible to use with distorded guitar, and time warp is part of the problem.
Why ? actually I am not sure that values used in the pool display for each file are coming from the audio wav file stream. In Windows (don’t know for Apple), there are as many as you want sub flows of data by physical file. This is already used by Window to store complementary infos and has been exploited to store ID3 tags