> This is starting to feel more like a hunt for syllables than an actual evaluation of the song.
No, I think that is a wrong impression that came across. I had rated a handful of Askus' songs over the past few years with 1 star as I noticed missing syllable separation along the way, but only now added a comment about it as he asked for such comments when giving bad ratings. So I was NOT hunting for missing syllable splits, I only added comments for previous ratings where they were missing.
> I’m not saying feedback isn’t important, but repeatedly hitting him with 1 point deductions over this feels unnecessary and, frankly, a bit rude.
Again, I think that this is a wrong impression. I am not "repeatedly" hitting on anybody, let alone Askus. We've been in touch via Discord for a quite a while. As I said, I only added comments to existing (but older) 1-star ratings because he explicitly asked for that.
But maybe let me explain a little more.
There are close to 30.000 songs here on USDB. Over the past few years, I have touched pretty much all of them, one way or another (adding meta tags, covers, audio samples, adapting GAP and or BPM, fixing text encoding issues, lyrics errors and the like). What I can't do is test every song in detail for playability. And since I am unskilled at judging pitches, I usually refrain from giving a rating anyways. However, what I *can* judge by looking at the song file is if words are properly split into the basic building blocks for singing - syllables.
Syllables are the fundamental rhythmic units of spoken and sung language, forming the building blocks of words. Each syllable contains a single vowel sound (the nucleus) and may include surrounding consonants, creating a distinct "beat" in speech or music. In spoken language, syllables help break down complex words for easier pronunciation and comprehension. In sung language, syllables are crucial for aligning words with musical notes — often one syllable corresponds to one note, ensuring clear articulation and rhythm. The structure of syllables, including their onset (beginning consonants), nucleus (vowel), and coda (ending consonants), shapes how words are produced and perceived. Recognizing syllables aids in reading, spelling, and understanding stress patterns, which are essential for natural speech and expressive singing.
In my ~20 years of being around UltraStar and karaoke, I think I can safely say that splitting words into syllables is consensus and has proven helpful for both creators and singers. There may be exceptions to this, e.g. for fast rap notes (no pitching anyway), or for swallowed syllables (syllables that are not pronounced, but are there in the written word). I see that it might be debatable to not split syllables of the same pitch ("why bother, it's the same pitch anyways" ), but even in that case I'd argue it will help both the creator and the singer to split these syllables anyways (allowing for a short gap between syllables).
The rating system is debatable anyways, since many, many ratings come in the form of "Oh, I absolutely love this song - 5 stars!", disregarding the quality entirely (many of the existing ratings are of that form). On the other extreme, people give 1-star ratings to perfect songs just because they were not able to adjust the GAP properly. Adding to that, some people forgot/forget to set the rating before the hit the rate button, resulting in a 1-star rating by accident even if they wanted to rate higher. So in the end, I'd say that a rating is only really helpful if it's an average of many ratings, so that it reflects some swarm intelligence. If 20 people give 5 stars, and 1 gives 1 star, you can assume it's high quality. On the other hand, if 20 people give it 1 star and only 1 gives it a five star, you can most likely assume that it's not high quality and that the one 5-star rating is by a fan that rated the song by how much they like the song itself.
That said, I understand that a 1-star-rating for a "perfectly singable" songs comes across as quite harsh. And indeed the least of my intentions is to push song creators away, discourage or offend them. Indeed, it's the opposite - I put a lot of effort, time and energy into USDB and tools around it to make it an enjoyable and fun experience for as many people as possible. And without songs, USDB, USDX and Co would be nothing, so I do appreciate every single song creator, taking the (considerable!) amount of free time to create songs and share them with the community. I have created quite a few songs myself over the years and know how much work that is and how much time it requires.
To sum up: I firmly believe that splitting words into proper syllables is (almost) never a bad idea (with very few exceptions). At least UltraStar Creator and YASS Reloaded offer syllable splitting functionality (within limits, e.g. slang words, proper nouns, ...), so this isn't really adding to the complexity of song creation either.
I apologize if my ratings or comments felt rude. Take my ratings as a quick win with an easy fix, and take it as one of hopefully many ratings. If the song is perfectly singable, feel free to give it a 5-star rating and I won’t argue against that. That makes it a 3-star rating on average, and if others join in on your rating, the influence of my 1-star rating will be negligible. That's how ratings work on Amazon or elsewhere - you trust the ones with many votes, not a single rating by one person.
> We already only have one or two people who make Swedish songs, and Askus is one of them. His songs are honestly excellent compared to many others, and I would hate to see him stop making Swedish content because of repeated harsh scoring over minor stylistic choices.
I would hate that, too! I just disagree with you that it's just a minor stylistic choice. Stylistic choices would be, at least in my opinion, the integration or omission of punctuation marks, the choice of quotation marks, the choice of where to split lines, the implementation of onomatopoeia, the placement of golden notes, the inclusion of spoken parts as freestyle notes, etc.
This is starting to feel more like a hunt for syllables than an actual evaluation of the song.
The song Askus made is already separated into syllables to a reasonable extent. The tone differences are clearly there, and in many places a ~ is used instead of splitting the word, which still works perfectly fine for singing. From a singer’s perspective, the song is very playable and flows well.
Giving a 1 point deduction for this feels way too harsh. The song is perfectly singable, and nothing in the singing experience suffers because of how the syllables are handled. If anything, being overly strict about this risks punishing songs that actually work well in practice.
I really think we should calm down with handing out 1 point deductions for things like this. Otherwise, there is a real risk of discouraging people who are genuinely doing a great job.
We already only have one or two people who make Swedish songs, and Askus is one of them. His songs are honestly excellent compared to many others, and I would hate to see him stop making Swedish content because of repeated harsh scoring over minor stylistic choices.
I’m not saying feedback isn’t important, but repeatedly hitting him with 1 point deductions over this feels unnecessary and, frankly, a bit rude.
