IMO it's a mistake to shoot for commercial levels early on in the learning curve. Better to be happy getting 90% of the way there and concentrate on all the other facets of recording. Squeezing those last few decibels of loudness has ruined many an amateur recording.
Start by comparing your recordings to commercial products that are in a similar vein. You can even import them directly into your own project to have as a handy reference any time you need them.
Just keep in mind that you'll probably never get yours quite as hot, but also that it doesn't really matter. If your record peaks at -1db instead of 0db, nobody is going to notice!
Perceived volume is determined by and measured in average RMS. SONAR doesn't have a tool for measuring average RMS over an arbitrary section or for an entire song, so you'll need an external tool for that. I use Adobe Audition, others use Sound Forge, but I'm sure one of the freebie editors will do the same thing, probably just with fewer bells & whistles. Here's what AA tells me about your song:
First, I look at the waveform. I can usually tell just by looking whether I'm in the loudness ballpark or not.
Looking at the waveform, you can see two distinct horizontal regions: the dense part in the center, and the thinner "hairs" jutting out of it. It is the ratio between these two that determines the balance between microdynamics and overall loudness. If the ratio is too low, that is, the "hairs" don't stick up high enough above the average, you lose "sparkle": the perception of transients like cymbal hits, as well as the sense of spaciousness and stereo width. Some genres are OK with that, of course, but given your genre I'd say you've struck a good balance: ride bell hits are crisp, but there is an overall meatiness.
Of course, this is all somewhat subjective. You can, however, get a much more objective reading by analyzing the average RMS value.
This is not as straightforward as you might think. RMS technically measures the power of a single cycle of a sine wave, which is fine for electricians but quite irrelevant to music. When we talk about average RMS in music, we mean an average of averages over time. It's the kind of calculation that would be impossible without the help of a computer.
What's important to remember is that average RMS is measured over a specific interval. SONAR's RMS meters use, IIRC, a 40ms or 50ms window by default. This lets you take small snapshots and see the RMS value as it changes moment to moment. But you can also measure average RMS over longer periods, up to the entire length of the song.
Here's what AA's "amplitude statistics" feature tells me about your song:
Bear in mind that this represents an analysis of the entire song. It's important to also measure this across just the main body of the song. With the intro excluded, the average RMS is ~-13db. You don't want intros and outros to skew the final number.
The average RMS of -13db falls into a good range for this type of music. It happens to be about where I try to land with my own stuff, which is of a similar genre. It's much quieter than your average contemporary commercial pop/rock product, which might between -8db and -4db average RMS (!). At those hotter levels, a great deal of detail and subtlety is lost, so -12 to -14 is, IMO, a better target. Especially for instrumentals.
I also see that you've normalized the file right up to 0db. I recommend that serious amateurs don't try to go for 0db, but to leave at 1 or 2 db headroom. Note that AA counted 5 overs - not enough to worry about, but still an indication that you may be a bit too aggressive. Pushing it right up to 0db exposes you to the possibility of intersample clipping. AA interpolates intersample values, so if I zoom in on one of the overs, I can see that your file will indeed exceed 0db when the analog waveform is reconstructed. It's too brief to be audible, but it's a canary-in-the-coalmine indicator that tips you off when you're perilously close to the edge. The effect is exacerbated by MP3 encoding, which tends to cause unexpected peaks due to ringing in the very steep filters employed during the encoding/decoding process.
The only criticism I have of this particular song isn't technical, but more subjective: its lack of
macrodynamics. After the intro, it hits the main level and stays there for the duration of the song. It would benefit, I think, from some targeted thinning. I'm talking about arrangement, not limiting. The microdynamics are quite good and I would not want to see it compressed any further.