I should probably elaborate a little more on DSP-based normalizing vs. adjusting to the same level via clip gain. Sonar has three convenient features when you change either gain or normalization:

• There's a short crossfade between the modified and unmodified sections of the clip
• The setting persists in the project until changed
• The waveform is redrawn to reflect these changes

So to level a phrase or word, it's just click + drag + hot key normalize. With clip gain you usually have to create four points, possibly tweak to set a slope if the change is too abrupt, and verify that you've hit the right level whereas normalize will always hit whatever peak you specify. As I find this faster and more accurate than adjusting clip gain, I never felt the need for clip gain to change the waveform - applying normalization or gain redraws the waveform anyway to reflect the change.
 
Some people might not like that this kind of edit is destructive, but realistically with the 64-bit engine you can change something as many times as you like without affecting the sound quality. And I do prefer these changes be "baked into" the audio.

Jeff Evans
Sorry Craig but you are wrong on this one.
 I agree that it is a personal preference...
 

 
It cannot be wrong and a personal preference at the same time. So Craig is not wrong. I agree 100% with his reasoning and thankfully so does Cakewalk.
 

Jeff Evans
They cannot all be wrong.

 
You are quite correct, SONAR has it right so they are not all wrong.
 

As many know around here I am a fan of VU meters too. The VU's in conjunction with the clip gain thing is a great combo. It is terrific for voice overs too. (or any track for that matter)  It is cool to be able to cut a longer part into smaller sections and tweak clip gains by eye on the fly. (Still faster than any other method and you dont need melodyne to do it either) I use the VU's to check and they let me know if something is still falling short or a fraction over so then you fine tune.
 
It is very easy and fast in Studio One to re render the clip gain adjustments into the clips as well. (one button click) It is great to see the waveform changes and to those who may be wondering if it is or not, the simple answer is yes it is.  In fact you don't have to render anything either you can just leave it all alone.  The waveforms look even and that is what you hear.  As I said before it is easy to read out how much clip gain has been adjusted.
 
Cakewalk got this a little wrong in my opinion. Once you start using it (especially a lot) like I say you won't go back.  Under pressure it is just so cool to section out a little quieter bit, add some clip gain until the waveforms match and move on. (Automation is eliminated now once those sort of things are done, see how it saves time)
 
If you track properly and even use some mild level conditioning on the way in then you should not need it either but in situations where levels do vary a bit (for what ever reason) it is very handy.
 
Are you sure you are just not making excuses for features that you would really like.  As we agree it is a personal thing and for me I think it is great and a wise decision. We are all different.
 

Jeff Evans
It is cool to be able to cut a longer part into smaller sections and tweak clip gains by eye on the fly. (Still faster than any other method and you dont need melodyne to do it either)...It is very easy and fast in Studio One to re render the clip gain adjustments into the clips as well. (one button click)

 
I'm very familiar with this way of dealing with clips, as Sony Acid (among others) has it, and I use it all the time in Vegas to adjust opacity (and sometimes for audio). I'm also very familiar with Studio One Pro, having done workshops on it at PreSonus's request. For now, let's ignore the fact that Sonar's clip gain automation lets you create slopes, so you can fade up when a singer runs out of gas on a note without having to resort to automation, as SOP's clip envelopes can't do this. Instead, let's look at physically quantifiable reasons why Sonar's approach is faster.
 
Sonar DSP method: First, it's incorrect to say the operation in Studio One Pro only takes one click. It takes a click and a drag, and that can happen only after you've split the clip into individual pieces. 
 
Second, in either program, if you want to adjust levels to the same value you have to do the same number of clicks and drags to modify levels. The difference is that with Studio One you have to do an additional click for every split you want to make, while in Sonar you have to hit a QWERTY key for every region you want to process. The physical action of hitting a key takes less time because it's one motion; splitting a clip requires positioning a mouse and then clicking. But there's one other thing that slows you down in SOP. When you drag, you have to be relatively precise with your drag to obtain the desired level. With Sonar, you set the normalization or gain once, and then hit a QWERTY key to apply and move on.
 
If you don't want to normalize but want to adjust gain by some arbitrary amount every time you adjust, then you're back to using Sonar's clip gain. But in this case, what takes the time in either program is confirming that you've made the correct adjustment and it sounds okay. You would need to do that in either Sonar or SOP. Arguably, you could get in the ballpark faster by listening with your eyes instead of your ears, but at this point I'm pretty facile with moving clip gains the right amount of dB on the first try. In this one particular example of adjusting clip gain, using the Sony method could be faster over time. But given a choice between that or the Sonar options, I'd take the Sonar options because the most tedious work is always working on tiny little pieces of something, and Sonar is faster for that. 
 
There's also less after-the-fact overhead because the clip remains intact for other editing or moving. For example if you decide you want to use the clip FX bin, in Sonar you just have to insert it once. I haven't worked much in SOP since X3 but IIRC this point is moot because I don't recall SOP allowing for clip FX...I could be wrong.
 

It is great to see the waveform changes and to those who may be wondering if it is or not, the simple answer is yes it is.

 
Using DSP in Sonar redraws the waveform automatically to reflect any changes you've made. This includes any of the DSP menu options.
 
Melodyne Editor, Assistant, or Studio method: This is even easier because when you open the clip in Melodyne and choose the Percussive tool, Melodyne has already split the clip in all the places you would want it split. For vocals it even isolates plosives, esses, and fricatives most of the time and you can use the Split tool if it hasn't. That eliminates the Sonar DSP method's need to define regions, and SOP's need to create splits. You still click and drag for each section, so there's no real advantage there, except the visual interface gives you a cue as to the nature of the sound you're adjusting, and you can hear an abstract representation of it as you adjust. 
 

Under pressure it is just so cool to section out a little quieter bit, add some clip gain until the waveforms match and move on.

 
If you're only doing a clip at a time, that is faster. However for the "cut a longer section into smaller parts" editing you were referring to, such as leveling out vocals, guitar solos, rhythm parts, you name it, Sonar's options are faster. I've put a lot of research into finding the fastest solution because I have to do this kind of tweaking all the time for narration, and nothing I've found is faster (other than recording the take perfectly in the first place). When I'm doing narration that lasts 45 minutes or more, if I can save even just a few seconds on every tweak it really adds up.

I have a follow-up. If you don't want to use AutoHotKey to do normalization with one key, you can bind to the keyboard shortcut for Process | Effects | Normalize. So you'd hit that key, then enter.
 
Also, there's an even easier way to do the DSP normalization thingie that I just tried on a vocal as a test, and it worked like a champ. I don't know why I never thought of this before, but that's the beauty of these kinds of discussions - they get you thinking.
 
Use AudioSnap's "split beat into clips" function (although of course they're not beats in this context, they're words or phrases). Then just click on a clip, type key to normalize, click, key, click, key etc. and they're done.
 
The only thing that slows this down is if the transients don't get placed correctly and you have to move them. But because most vocals have spaces in between words and phrases they're pretty easy for AudioSnap's algorithm to catch.
 
Finally, V-Vocal has a dynamics tab where you can draw in rubber-band amplitude changes and V-Vocal redraws the waveform to reflect the changes you've made. Because you're just dealing with amplitude, you don't get the dreaded "V-Vocal phasey sound."
 
Come to think of it, describing how to do all these things would make an excellent column for Sound on Sound!

I was disappointed to find that I cannot select a small section of a clip and use the same technique to adjust only the selected sections gain. :-(
 
That makes deessing that much more difficult. Unless it works that way with the clips gain shown. I'll check later.
 
BTW - Thanks Jeff for the tip about using a meter. That made adjusting the gain more informative.

@Guitarpima
 
If you talking clip gain adjustments than you can select a small section by dragging in the timeline and then adjusting the gain with the mouse at the top of the selection. Hope I'm commenting on what you're mentioning.

Guitarpima
I was disappointed to find that I cannot select a small section of a clip and use the same technique to adjust only the selected sections gain. :-(
 
That makes deessing that much more difficult. Unless it works that way with the clips gain shown. I'll check later.

If you're talking about SOP, you have to split the clip to do that.
If you're talking about Sonar, you can draw the clip gain curve however you want, or use the click+drag+gain/normalize technique I mentioned on the clip itself.

I think the wave form only grows or shrinks in size when it is actually edited to do so, like through the Process - Gain command. Like all automation, the Gain Automation line doesn't change the actual wave form so a visual reference as you suggest could be confusing to some.