Best way to fix a fast guitar solo

Cindy . . . which version of Sonar do you have ? The newest versions have a feature (Melodyne Essential) which can do some "re-tempoing" (probably not a real word, but you might get what I mean ?)

You could try splitting into heaps of clips and using vocal sync. I havn't tried it, but always envisaged a use for it like your describing.

Hey Boydie. Do you use Melodyne (or anyone else reading this thread) for this type manipulation yourself? I find for anything other than the most simple notes of guitar work, that the process destroys lead in/out of individual notes. Great for vocals. And bass sometimes. But anything intricate like finger picking, moderately fast solo's, fast rhythm etc, that it just doesn't work at all. Manual correction of individual notes, sure, but not bulk quantisation like your suggesting. Please don't think I'm dispelling your suggestion. I really am curious. Is it just me?

Oops. I did not see Kalle's post. He is describing the same process as I am. It works well.
 

Soundwise
Check this tip (and the thread)
http://forum.cakewalk.com...074655-p2.aspx#3110584

In theory that would work but there are a couple problems...
 
One is that you end up stretching material that doesn't need to be stretched (if the song is mostly in time as the OP says).
 
The second is, in that article Craig neglects to mention all the transient marker setup work that needs to be done before snapping in bulk like that. Most of time, unless the material is super clean and has strong defined transients you have to go through all the markers by hand to reposition them at the actual transients. Transient detection is a little better in Sonar these days (and Audiosnap in general is working WAY better than it used to) but it's no where near perfect. On a fully mixed stereo file there would have to be a LOT of prep work making sure the transient markers are actually  at the transients as well as disabling/inserting transients that are false positives or missing.
 
I literally just did a pile of this over the past couple of days on an extremely simple, dry DI guitar track that was just a steady chug. Even with that it was extremely time consuming getting the markers set up properly.
 
Best to avoid that type of AS work until it's absolutely needed which is why I mentioned the Split/Slip Stretch method to ballpark it then Audiosnap to do any extra cleanup needed.
 
Not that your suggestion isn't good and will be worth it for the OP to check out. I just think it's not quite as simple as Craig made it seem and if most of the material is in time isn't necessary.
 
Cheers.

Beepster

Soundwise
Check this tip (and the thread)
http://forum.cakewalk.com...074655-p2.aspx#3110584

In theory that would work but there are a couple problems...
 
One is that you end up stretching material that doesn't need to be stretched (if the song is mostly in time as the OP says).

 
Based on the OP, I think only the track with the guitar solo needs to be fixed.
 

The second is, in that article Craig neglects to mention all the transient marker setup work that needs to be done before snapping in bulk like that. Most of time, unless the material is super clean and has strong defined transients you have to go through all the markers by hand to reposition them at the actual transients.

 
When working with a single guitar track the transients usually are well defined (especially if the track is dry and you're going through an amp sim), so they don't require a lot of manual labor. 

Hi, Craig. I actually read the OP about five minutes after waking up and yeah... blurred on the exact nature of the task. Also wasn't taking a swipe or anything... just pointing out to the OP (who may not have known otherwise) that AS transient detection needs to be double checked at the very least.
 
Thing is though, in regards to the detection on even a guitar track (especially if there are already effects on the track) IME the Markers need to be corrected before bulk stretching is done.
 
Like I said I just went through this with a current project. The guits were DI and picked at every 8th note (or something like that) conistently throughout almost the entire song. The pick attack transients were quite clear and pronounced and there was no ring out (palm muted chug). The original performance was actually almost spot on with the timeline but the client want it to be robotic (industrial) so I went the audiosnap route to do so.
 
As expected and often mentioned in tutorials (yours included where I've picked up a lot of my AS knowledge) there were quite a few Markers that were slightly off the actual transients, some were unwanted and doubled up (due to maybe slight string noise) and some were missing (even though the transients were clear and defined).
 
Just the nature of detection stuff like this and cleaning things up properly before stretching, as you know, is needed beforehand or you get unpredictable results. Nature of the beast.
 
For groove extraction eq filtering or gating/transient shaping can be used on the source material to make the transients stick out more and THEN the detection works great (and I actually had to do that on this project as well)... but you can't do that on material that is being corrected because then you just ruined the sound and can't get it back after the bouncing.
 
You of course know all this and perhaps you have some tricks or your equipment somehow gives you better results with this type of thing than I get but I just didn't want a new user to look at that and think it's an totally automatic process. Never is.
 
However since this is just a backer track that is being used to build around and actually the OP is a little vague about exactly what is being attempted (like is the solo/guitar track the only thing that speeds up and all the other instruments stay in time? Is this a multi track file where all tracks need adjusting or a stereo mixdown? They mention vocals and there is an implication that AS adjustment throw everything else off, etc).
 
In any case, if it's just a scratch track then doing a clip stretch instead of audiosnapping still seems like the better/easier route to record over top of. I've done it quite a few times before when collaborating on covers. I'm even about to restore one of my old band sessions using these principles (but WAY more in depth and complicated stuff... essentially correct drums and vocals then rerecording everything I can on my own).
 
Aaaaaand to top that all off audiosnapping a solo (compared to the simple chug I was talking about earlier or a drum beat) is pretty finicky at the best of times for transient detection and correction. It's gonna be a lot of notes flying around in weird ways with different textures and intensities that would need to have the markers combed through to make sure they are all there and in the right place. THEN to do a bulk correction would still be hit or miss unless it's a super simple and rhythmic solo because we all know what happens when you try to snap something like that in bulk. Some notes go one way and others the other way. Things that were in time go out of time. Things that were out of time and end up in time might sound weird. Groove and feel gets lost and it's a TON of work for... well not very good results.
 
Doing it note by note by hand is the way I would deal with stuff like that but it ain't worth it for a backer track.
 
Totally not dissing or being argumentative of course. Just putting my real life experiences with this stuff up in the forum archives for those who may search this topic so they can know what to expect and how to solve the issues that may come up.
 
You know I lurvs ya, Craigo... and really a lot of this knowledge has come from your own tutorials.
 
Cheers!

Sorry for fan girling, but since Craig Anderton is on this thread, I need to point out that this is all his fault!
 

"I've been adjusting the audio transients, but the back half of the song needs to be moved and a lot of adjustments need to be made"
 
That's the part that's leading me to think the solo threw the rest of the track (all instruments) out of whack.
 
I've been in the EXACT same situation before (a few years back when I was just getting started) and attempted the EXACT same method for correcting it. It turns into a complicated mess with Audiosnap. Especially back then before they fixed AS (it was unuseable before... I am very happy it's working better now).
 
In fact that is the very same project I referred to earlier that I am about to tackle again using the methods I described after a bunch of experimentation so this has been on my mind lately. Kind of why I'm here and mouthing off.
 
Well that and I saw "Fast Guitar" and just hadz to click. lol
 
;-)

So that stretching thing will work.
 
However, just for your own knowledge, you can also take that off time clip and alter your timeline to it (so you keep the off timing and record around it).
 
This can be done with Audiosnap using the Groove extraction method but for a short section (and because I'm anal and like to do things manually)... check this out.
 
Right Click on the Time Ruler and select "Set Measure/Beat At Now". A window will pop up asking you what Beat/Measure you want to adjust (this will occur wherever the Now Time line is currently at).
 
Now that you're aware of this little hidden feature you can set your Now Time at beat points from your HUMAN performance and translate them to the Timeline manually.
 
So in this case you'd go to your solo section and find the Measures/Beats on the performance, set your Now Time at those points (one at a time) doing that "Set Measure/Beat" action. Do that throughotu the project as needed.
 
Now your project tempo will match the raw human performance perfectly (if you did it right). Your Metronome and any MIDI stuff will land right where it should and playback in time.
 
As I said Audiosnap can do this type of extraction a little more automatically but that requires you checking where the auto detected transients appear and doing a bit of cleanup and mental calculation so it all lands properly.
 
Just rambling about cool stuff you can do that seems like it might be useful/helpful/interesting to ya.
 
Cheers.

I thought I understood this was a single guitar track that also has a solo part in it where it is sped up? If so I'd love to throw it into Melodyne and see if I can fix it. Should be a 30 second operation if it works as advertised. 

As you move forward and learn tools, workflows, etc., do not lose sight of the validity of a simple re-tracking. From the perspective of time, playing for 3 minutes to edit for 3-hours can sometimes be one's nemesis; when playing for 20 minutes may be a better solution - plus it makes tracking go more smoothly in each iteration.
 
Reason I say this is I have a friend who still records to tape, so he has tracking down to a fine art. I continually harp on him as he moves into the "DAW world" not to get enticed by all the editing tools that are available.

Sanderxpander
If it's just that single guitar track, Melodyne should be able to get a good tempo out of it which you can then straighten out. 

It's actually a "live" track of one guitar and two vocals, only part of which is the solo. After looking more closely, I've found other time issues throughout the song. Melodyne is a tool that's high on my list of things to learn.