The problem with me and singing is that I’m am absolutely terrible at it lol!
Well, plenty of people don’t like the way their own voice sounds. Even worst case and you’re correct, are you just not pleased with the sound of your voice, or is your pitch just all over the place? Singing phrases in the context of transcribing only requires decent pitch 
1 Like
My pitch is all over the place. terrible!
I have been thinking about transcribing, and I think a good way to brush up on skills, and to practice skills, is to go for jazz players. Very little string bending and their phrasing is a little more straightforward than rock players.
Oh, hey! Only just seeing this now.
I highly recommend soundslice to simplify workflow. One window for audio + notation. Small learning curve. Corey just put up a new demo here . You can see videos of me transcribing with it here and here.
And more misc. transcriptions at my soundslice channel, here.
(This is assuming you want to document/write down what you’re transcribing. If you’re not going to write anything down, soundslice’s advantage isn’t as significant.)
I could comment on this just that it depends on what your goal is with the transcription. I’d say in general it’s good to be able to notate things, but the priority level depends on a lot of factors. For whatever it’s worth, when I’m struggling with something (especially rhythmically) I like to write it down so that I don’t have to just keep hoping I’ll remember.
I guess I could argue there’s a push pull because let’s say for the same quantity of music, the more you write down, the less you have to use your ears, but the faster you’ll produce a finished, playable product, and also the more accurate the transcription will be. The less you write down, the more you have to internalize the vocabulary/concepts (which is a good thing) but if the goal is simply to be able to play verbatim the thing you’re listening to, in the end this could take a lot more time and be a lot less accurate, assuming the vocabulary in the clip is stuff that’s challenging for you.
Do yous use headphones when listening to the tracks?
Pretty much always. If there’s a very clear mix, and/or I’m only trying to figure out an instrument that’s a higher register, I might just use the speakers but pretty much never laptop speakers or iphone speakers unless it’s like a tin whistle or something.
Do you not have your guitar with you, and instead use a midi piano?
Basically neither. I’ll often have my guitar on my lap but find myself only using it if I’m not confident in something. Even then, since I’m inputting notation into soundslice, I usually have it set so that as I type in a note, the computer plays that note back. So if i enter the wrong note I’ll hear it right away.
Now, that part is more in the long-term skill department - through a combination of things I’ve worked on over the years, I’ve gotten pretty quick with simply just knowing what something is on a first listen. For a melodic line (one note at a time,) hearing groups of 3-4 notes is pretty simple and I usually can identify most sets/sequences on a first listen unless it’s something more dense or complicated.
But when you’re newer to all this and not confident you’re doing things correctly, I think it’s fine to check stuff on guitar, sing it, engage the midi playback, whatever, to be sure you’re hearing what you think you’re hearing.
With transcribing, it’s super easy to make mistakes and not realize you’re making mistakes. To that point, it’s funny that in that old Wes video I linked to earlier in this post, I actually got the very first note wrong. I just got it in my head that he was starting F to G and didn’t really check it enough times, and for some reason throughout the vid I just accepted that that was the note and it didn’t catch my attention until later.
Since I’m often transcribing stuff for lessons, and often teaching the same tunes, I’m very frequently forced to review transcriptions I’ve made and I’ll typically hear new things every time, or mistakes from earlier versions. Consider that, if I’m a “good” transcriber, and still able to find mistakes in my own transcriptions each time I go over them (whether they’re from yesterday or from 15 years ago) then when you’re new to it, there’s definitely just going to be tons of stuff you’re not going to be hearing correctly, tons of details you’ll inevitibly miss, etc. It’s just part of the nature of the whole thing; we can hear what we can hear. And a transcription is really only “done” when you just say “ok f*** it, this is probably good enough” because you can always go over it and over it and over it…
It’s up to you how accurate and detailed you want to be, for whatever purposes you’re getting into transcription.
By the way, a tip on building accuracy and building your ears in general: If you find there is a certain type of thing that’s much harder for you to transcribe accurately (maybe chromaticism, maybe larger chords, maybe lines with bigger intervals, maybe funk rhythms, etc) I think it’s useful to isolate that skill with some sort of more formal/structured ear training, programs like iwasdoingallright.com can be very good and useful.
Example: the first time I ever tried transcribing jazz guitar voicings I had absolutely no idea what I was listening for and basically had zero confidence in what 75% on the notes were. But at that time I hadn’t done much ear training with more basic things like triad inversions and other more basic 3 note chord structures, so hearing these rapidly moving, complex 4 note chords voiced with lots of slides, it was a bit futile. I needed more ammunition! Um, aural ammunication? (Band name?)
Do you run your audio files through any audio seperation software such as Spleeter, to try and extract the guitar tracks?
No but only because I’m not really familiar, haven’t dabbled. I do often use EQ and panning to try to dig into what I want more.
Hope that helps!
2 Likes
@JakeEstner Jake, amazing post, thank you so much!
One question - how do you deal with fast played lines, where the note are pretty squashed up together?
Just watching your Levi Clay challenge and you are fast at this!
Hmm, I’m not 100% sure what you mean, but I can say I’ll adjust the speed of the playback for sure, which you can do on soundslice and transcribe, and I’ll do as few notes at a time as I need to to get it. For example, whether it’s something super fast or super slow, I’ll adjust playback speed and try to grab just a few notes at a time. The more familiar the sequences are, the more notes I can usually do at once, and conversely of course the more dense and complex it is, the fewer notes I’ll do at once. So in some cases I’m going one note at a time.
The only problem I run into with speed is sometimes it’s fast enough that slow down features garble the pitches a bit, which I guess is probably less of an issue with higher fidelity recordings.
1 Like
Bumping this after 2 years!
I’m working more on chords. Are there any general tips when it comes to transcribing chords?
I’ve been listening to a lot of kpop as I am fascinated with how well some of the songs are composed. There seem to be modulations and chord substutions in a modern pop context. Really good stuff.
Here’s an example:
For a start, I cheat and load the song into either Sonic Visualiser or download from Youtube and put it through spleeter to isolate the backing and the bass. It’s quite hard to work out some of the voicings.
Here is what chordify says -
Here are the chords I have:
Any hints or tips would be greatly appreciated.
No such thing as cheating
The only advice I can give is to listen to the overall chord “quality” (maj/min/dom7/dim/aug) then listen to the bass to identify the inversion. Hearing inner voices was always a challenge for me. Not sure if you are looking at that level of accuracy (i.e. every single pitch present in exact order)
1 Like
I too am on this journey so I don’t really have any definitive answers for you. My dad can figure out the chords on the fly, as people sing songs he’s never heard before, and accompany people live that way. I’ve always wanted to learn that, but it’s hard work.
He picked the skill up listening to the radio as a teenager. Whenever a Beatles song or whatnot came up in the rotation he would listen his heart out and try to reproduce the same noises on the guitar. Eventually he got damn good at recognizing them noises
Two books have stood out to me, and I think you might enjoy them too:
- The jazz harmony book by Berkman. The book is on the theory side of things, but takes a very practical approach. It starts out with the I IV V chords and goes on the re-harmonize in all kinds of ways to show you why those more exotic chords make sense and how they come into play. It comes with audio so you can listen along which is obviously super helpful.
- The real easy ear training book by Roberta Radley This book is very hands on and shows you various strategies you can use to narrow down the list of potential chords and how you can hear them. This book also comes with audio, but it also has an extensive list of suitable music that you can use to practice the various skills on. That list was worth the price of the book alone to me. The title makes the book sound stupid (and nothing about this was easy lol), but the book is a real gem imo.
I also used EarMaster a lot. It’s a dated piece of shit in many ways, but the custom exercises let you practice recognizing chords in a very efficient manner. Depending on your level you can start with hearing major/minor, then identify all the various 7th chords by name, then you can progress to all the 9th chords etc
Once you can hear the quality of the chord and can pick out what the bass player is playing (usually the root of the chord), the rest falls into place quite quickly.
1 Like
chords and chord progressions, two perspectives that overlap significantly:
-
Brute force:
a. identify the bass note/lowest voice
b. identify the highest voice you can hear in the chord
c. try to fill in the blanks of inner voices - some tips on that:
c1. set things up so you can loop/isolate an individual chord
c2. try some “yes or no” testing with individual pitches. For example…ok I think I hear a 7th in the middle there, let’s play a 7th (the individual tone) then listen to the chord and see if I can hear that tone. Ok maybe, let’s compare it against b7, 6, or root…hmm actually I definitely hear the 6 in there, no 7. Time consuming? Sure, but gets faster the more you do it. And also a good way to really check accuracy. Chords are groups of notes - of all 12 notes * X octaves, every note is either a “yes” or a “no”, basically (without getting into overtones, which can be very annoying when trying to figure out chords) -
The more you know about harmony, the better, the more you can sing common cadences and progressions - and think and hear in roman numerals - the better. When you understand harmonic movements common in different genres, you’re not hearing chords anymore but movement. It takes time but imo has advantages well beyond just being able to transcribe. For ‘jazz related’ progressions I make sure students are familiar with the following the triads and seventh chords of the major and minor scales, secondary dominants, backdoor cadences, tritone substitutions. I make them be able to both analyze and transpose relatively simple pop tunes (think mainly diatonic triads) as well as basic “Real Book” tunes like Autumn Leaves, there Will never be another you, Alone Together, etc.
When you study harmony, things are much easier to recognize! The ‘brute force’ method becomes more so a means of double checking your guesses. However there’s an efficiency question here that becomes a very personal/individualized thing: you can spend a year studying harmony to make it faster to figure out songs, or you can spend the year figuring out songs. How much to focus on one vs the other depends a lot on the individual, imo.
Hope that helps!
2 Likes
That’s a really good point. It is like the “chunking” concept. Don’t think in letters or even words but sentences/phrases.
1 Like
Thanks Joe. With regards to bass notes I find that if I shift the track up an octave it’s easier to hear the bass notes. No idea why this works, but it does!! I’m not after 100% accuracy, just close enough is good for me.
I was looking at the EarMaster site and am wondering if you need to be able to read standard notation for it to be useful? I can read standard, but I am very slow at it!
Thanks Jake. I like this brute force method. I had been working off bass notes, but hadn’t thought about identifying the highest voice in the chord - that seems to open up more possibilities.
No need to be able to read standard unless you want to get into sight-singing If you were to do the chord quality quiz the app will play a chord and you’d hit one of several buttons marked e.g. Maj7 or Min7 to submit your answer.
1 Like
I’ve transcribed a lot, but I don’t use anything complex. I record the clip on Audacity and then reduce the tempo to make it easier on myself. Audacity has a good algorithm for tempo change without introducing undue distortion. If the instrument is on the left or right side I’ll take just that mono side to help isolate it, but that’s all I do for isolation. I have my instrument with me because I need to see how the transcription feels under my fingers while I’m figuring it out. I’ll tab it out while I’m transcribing using an app (I’ve used Tabledit for over 20 years now) . I’ve not used headphones before, just a good set of speakers, but that’s not a bad idea.
This has always worked well for me.
2 Likes
If you’re on Mac/iOS, I’ve found Anytune+ invaluable.
Besides the software, I’ll search for isolated/STEM tracks and live performances or covers for neck positioning, pick direction, or see someone play the dang thing. Songsterr isn’t bad for finding things. But like most tabs - it’s never 100% according to my ear.
The rest of this is a lot longer than expected. It’s not a rant, so bear with me. I doubt I am, but I’m also hoping I’m not alone in my opinion.
I understand that “close” may be good enough if you’re learning a song for a one-off and, or short on time, OK. If you are intent on understanding the feel of a player, slowing things down and getting it as close as you can becomes very relevant.
Not because “That’s exactly how they did it.” - but because it will highlight rhythmic nuances you can’t hear at speed. The space between the notes. From a Cracking the Code perspective, one note/nuance makes all the difference. An example is Yngwie using pull-offs when descending. Troy thought he picked everything until he slowed it down and saw this little nuance that changed things significantly.
One may call this obsessive as if it’s unhealthy in some way. It would be if there weren’t a real point to it, and it’s not bragging rights for playing note-for-note.
Slow Eruption down, and what sounds like pure blasting is a metric ton of swing feel. EVH never played purely ‘straight’. There’s always a pulse in there. That’s one reason his stuff sounds playful, and your body wants to groove.
The same is true for Eric Johnson. Subtle stuff that, at full speed, sounds like a specific note value, but slow it down, and you find a drag here and a rushed note there. Miss those notes and the rhythm - and therefore the feel - changes. It’s slight, but so is a fingerprint.
Alex Lifeson’s solo in YYZ is another fantastic example. Typical of his playing, it’s a very angular, oddly phrased, almost ‘out’ sounding solo. It sounds like he’s blazing notes (and he is). But doggone if he hasn’t woven a deep groove throughout the entire thing. When I discovered (heard) it, I thought, “That’s crazy…and now the whole thing makes SO much more sense.”
We obsess about pick techniques, vintage vs. new, and what amp/speaker/cabinet/strings/pedal/pick the player used.
Why not this? What if Troy hadn’t seen things like Yngwie’s descending pull-off trick? Would we even be having this conversation?
If you made it this far, thank you for playing. I hope this makes some sense.
1 Like
Exactly. To me, the stuff I’m into emulating (namely EJ) is either playable vs not-possible if you don’t have the exact notes and even fingerings correct. Obviously the picking has to make sense with the mechanic but there are often several ways that would “work” to get the correct notes. It’s not until you get it to the target tempo that you’ll discover which “correct” way is humanly possible, because some of his most challenging runs could take decades of failure instead of just a few weeks of doing it an easier way (i.e. EJ’s way).
That’s kind of a long winded way of saying, if you don’t spend the time to get it “right” you may end up with something you are anyone might not be able to actually play lol
That said, as of late I’m more on this kick of just doing things “my” correct way. In other words, whatever makes the phrase feel easiest for me to play. That could involve changing actual notes in a solo of someone else’s I’m learning. That’s obviously out of the scope of this thread, but assuming part of the reason someone would transcribe is so they could “play” the challenging music they like. An adaptation of the original could be an alternative.
1 Like
Understood.
That’s why I qualified the intent to be on understanding the feel of a player.
The one thing I want to emphasize is that my comments aren’t pointed at the right notes, fingerings, pick technique etc.
It’s about the one thing almost no one discusses, and that’s the micro-timing a player has and how that affects how everything feels.
In the context of understanding a player’s nuances, yeah…you can get ‘close’ and do your own thing. But then you still really haven’t sorted out what they did. The YYZ solo taught me so much about Alex, it’s silly.
My point with the nuanced timing is to dig into a player’s ‘feel’. The notes, fingerings etc. are fine. With enough focused practice, almost anyone can figure out where and how to put their fingers in the right places. You can find those players all over YouTube. Flawless technique, with apparently no effort…but its lacking somehow.
I don’t think the individual sense of rhythm gets discussed enough in guitar circles. We talk about it at a macro level but tend to dismiss most everything else it seems.
1 Like
Ah sorry I see I missed your point. Yes, that’s all vital also. For some players, like the ones you mentioned. Probably not as big a deal in learning someone like Al Di Meola since I’d imagine his phrases are pretty much locked to the grid.
When I’m in my EJ copying phases, I’ll regularly loop the phrase I’m working on in SoundSlice and set the tempo to like 65% and just focus on playing each not exactly when he plays it. I’ve found that’s an area where gradual (5% or 10%) increments in tempo are pretty helpful. On the good days when I get to 100% it really does feel like I’m playing with his vibe. It’s pretty hard.
I haven’t tried that with any Van Halen stuff, other than the “Right Now” solo but I recall similar (good) progress with that one. I’d always sort of written his solos off as things I’d never be able to play. The notes themselves and the speed weren’t quite the issue, it was nailing the nuances. Even at tempo it just didn’t sound like him.
1 Like
Lots of different ones:
Sibelius (subscription, but has a free tier)
Finale (not cheap)
MuseScore (Totally free, Win/Mac/Linux. No ads…works well)
2 Likes