Initial thoughts after playing one popular score[0] in FF dev 46.0a2 (3/2). I can only see four measures per screen, which is fine. However, by the time it scrolls to the next page it's already played 2.5-3 measures so really all I see is a constantly scrolling mass of notes, none of witch are the ones being played at the moment.
Playing another popular score[1] I get constant notifications of other people entering the score, which seems to slow down the animation a bit behind the audio. The slow scroll of the staves when the song moves to the next line also makes the animation a bit sluggish until it gets to the top and is smooth again (but starts scrolling about 1.5 measures later anyway).
It seems like a neat project but the playback usability leaves some to be desired.
When seeing the pagination, I imagined the app's author thinking it would make people right in their element. However, realistically, importing the annoyances of paper scores to the browser seems like a mistake. It makes more sense to have the score go infinitely towards the right, which they do on their non-default ?layout=track option.
As an aside, what about the legality of publishing copyrighted music that can be played without paying royalties?
We had and have so many debate over that topic. Actually, both ideas are true.
As you I would prefer the track view, but many had claimed the "page view".
One thing for sure, the scroll isn't convenient at all...
We're working to address that and improve the overall experience :)
As an electronic musicial (http://www.soundcloud.com/decklyn) I'll give you some feedback from years of hanging out in a DAW. I'll bring up my perspective on some points others have offered and a few other items primarily related to usability. Note that I've spent a lot of time in a piano roll rather than notation software so my experience is very colored but hopefully this is somewhat useful.
1) The need for instruction others have cited I think is a bit over-stated - I was able to get in and start writing immediately without any introduction.
2) Using social media for login eliminates the barrier to entry for me personally. I was fine signing in. Sure it might be fun to allow an ephemeral experience if people just want to check it out but I wouldn't put that at the top of your backlog.
3) Placing notes.
IMO placing a notes should be a toggle not a draw function. I don't want to select a note and delete it - I want to double click it and have it disappear as you would in pencil mode in a daw. At the same time, I don't want to click a note and then click on the length of the note to change it - I want to drag the selected note's length.
3) If I want to change a notes length, I'm forced to change a group of notes existing at the same time. I'm assuming this is a limitation of the modelling - eg that for each fragment of time, a group of notes exist together. This inherently limits me from producing music. If I want to write a quarter note A and two eigth notes C and then D I have to draw a twice and then tie it - this is a lot of work IMO. Again maybe it's because I'm used to the piano roll in a daw and I haven't worked in notation software as much but I fell like it should behave somewhat similarly - that notes are more of a toggle.
Using ties doesn't seem intuitive - although it lets me span the length of a single note, maybe you could have a tool that treats notes like a grid and lets me click a noted and drag it in the grid - vertically for pitch and horizontally for length.
Just some ideas based on things that I found where a bit difficult for me to get done having a quick trip through.
It's cool though - thanks for making this. I'd consider laying ideas out here on the go.
Hello there!
Thank you for your message.
I think that in the future, we will add more freedom to the edition, to get to something close to what you describe. For instance, we would have an abstract way of handling note duration, that would then be translated to one or many tied notes.
As someone who has used both a DAW and musical notation software in the past, it must be mentioned that they are two completely different paradigms. This product is designed for writing musical notation. A DAW is designed for manipulating the arrangement of samples in a multitrack timeline.
That's a very narrow description of a DAW. OP was specifically talking about the piano roll - a feature that obviously differs from traditional notation but is still very much related since both aim to achieve the same goal.
A DAW surely isn't limited to sample-based arrangements. In that light your description of a DAW isn't wrong, but very much unrelated to the point OP was trying to make.
This is great and an idea I've toyed around with building. I think we'll eventually see web-based music composition and DAWs in the browser take off.
As someone who's become used to composing with guitar tablature (a la Guitar Pro) instead of musical notation, it'd be great to see this mode supported as well. The fact that this has MIDI export built in is another great feature and a requirement to prevent lock-in. Good job.
I agree. High-end audio manipulation will remain native, but I think for a simplified use-case (composition, lower quality renders) the browser is a good sell.
That and the tooling is almost here with MIDI.js[1], Web Audio[2] and Web MIDI[3] as a working draft.
In fact, latency and audio quality (because of file sizes for sampling and because of available resources for synthesis) are two big limitations in a browser.
Also, we want to enjoy the capabilities of the MIDI output features, in order to route MIDI signals to Virtual Instruments on the computer (through virtual MIDI ports)
On mobile at least, it sucks (trying to force sign up, being presented with a page saying just 'unauthorized' on a link offered when you have not signed up). Also, from what I gathered, this is not an online daw or composition tool, is it?
Some thoughts: the typography is rudimentary (lots of spacing issues and collisions), the number of notational features in lacking (can't seem to find how to add a turn or Italian tempo markings, for starters). Note entry is inconvenient (although I didn't try using MIDI), not so happy about the TOS as far as rights to your own work is concerned. I also had a number of failed attempts at xml importing, not sure what eventually led to a sucess. The interface is also pretty simple and intuitive to use, especially compared to native apps.
Indeed there are still many notations lacking. We're doing our best to ship as fast as we can.
Regarding the import, music notation softwares are not sticking to the standard and it's a real pain to be "compliant" with them... :(
Yours finally worked due to a hot fix we did ;)
Regarding the ToS, we didn't think them this way but I completely understand that it can be a concern. Will update them within the upcoming days.
Give it a shot in the free Soundslice MusicXML viewer -- our importer is pretty damn good, despite the pretty horrific/inconsistent implementations of MusicXML export out there.
As someone who recently switched from academic music composition to a corporate job, this is really cool.
I've used Finale in its various versions for over 6 years now, and it's still widely considered in academia as the best notation software, not because of its ease-of-use but because of the depth to which one can add weird/special notation to the score.
In my experience, either composers are using Finale, or a combination of Sibelius and Adobe InDesign. This being said, most of the (albeit mostly American, or Americans in Europe) composers I have met/worked with often use notation software as the last step in the composition process.
The collaborative aspect of Flat.io is really nice, but until the depth of notation catches up to the desktop clients like Finale/Sibelius, I have a hard time seeing (at least) concert/academic composers make the switch.
Side note: is it possible to rebind the note values to different keys? I've gotten so used to Finale's 4=eighth, 5=quarter, command+(NUM) for duplets/triplets, etc.
I'm a hobby pianist with occasional paid gigs. I'd love to use this to play along with scores. Here is my feedback purely from a playback perspective. I haven't tried entering / editing any scores yet.
Some issues:
- Vertical page scrolling is not as intuitive. Could the pages be moved into place from the right? (horizontal scrolling).
- Scrolling is not at all smooth and just far too slow throughout (regardless of the mode used).
- No support for multiple systems per page. In other words, if the page could fit more measures, the next measures should be wrapped.
Feature requests:
- Consider allowing a no scroll mode - just skip to the next page.
- Also, allow configuring of page turns X beats ahead of the playback location. For example, I prefer folks to manually turn my pages 1-2 measures before the end of the page as I will have memorized the remaining measures.
- Could you provide an option for the score to resize to use the maximum available vertical space?
- Landscape page mode?
- Hide entire instruments (therefore reducing the system height).
- Support multiple systems per page (line wrapping of measures).
You can switch to a vertical mode with the button on the top right corner (||||). It should also mitigate the scrolling issue. Meanwhile we will do our best to increase performance on huge scores.
For your features requests:
-A no scroll mode will definitely come
-Yes, you're right, for now it's made for listening but not for playing we will soon add a "Performance" mode in order to smoothen the "page turns" and to do it automatically with the sound.
-That's a neat idea we will consider it
-It's already available in the display options (the second button on the top of the first page, on the right of the instruments button)
-In fact it's a good idea in order to have more space when working on a piece
-You can have multiple systems in a single page. I assume that you have many instruments. We try to keep instruments grouped, not to have instruments scattered through multiple pages when possible
This is a very impressive piece of work. It's addressing a similar problem to my friend Adrian's project www.soundslice.com - the complexity of hosting interactive sheet music online is enormous.
How is a user supposed to find out what this site really does without putting in an email address and agreeing to terms of service?
So many sites do this:
> Cool Title!
> Sign up here.
> Some Web 2.0-looking boxes with a few words about how cool the thing is.
> Sign up here.
With no obvious way to something that isn't either a sign-up box or a few more meaningless words.
And then way down at the bottom in tiny font I see "Music Score Editor." Ah-ha! I think, and click. No... that just scrolls me back up to those meaningless words.
Next to that is "Guitar Tabs." Now that one DOES lead to an example. There is a real example! With no way of guiding me to this link, in tiny font at the bottom, and under the previous meaningless link.
And finally I realize that there's a link to the top, "Popular", which is not a helpful word and is, again, next to a meaningless link ("Education", which leads you to more words).
First thing on the landing page, it says: "Write your music scores online"
Create, collaborate and discover sheet music with your web browser
I found the "popular" link pretty quickly. Click that, and if you STILL unclear on the site, then no matter what they do, you're unlikely to be happy...
Harsh, yes, but I had the same experience. I scrolled around looking for some actual music, then left the site to check HN news comments.
Thanks for the tip about the "popular" link, I had missed it. After checking out a couple scores I signed up, looks like a cool site. Had the site simply put their top 4-8 songs (with that nifty preview-on-hover) on their landing page, I imagine their sign-up rate would skyrocket.
I agree. If "Write your music scores online" is not enough for you to understand the product, then you are probably not a target user.
Similarly, lots of ad tech and dev tools landing pages are unintelligible to people who probably don't want to use the product, but perfectly reasonable to those who might.
I can't see how you can make such a comment while being efficient. When you're looking for some software, you don't want "X that does Y". There are already solutions (maybe hacky ones) to do Y, so you want some X that will be good in your workflow. And this is not something you can decide with "Do Y. Awesomely."
Honestly the current page is not so bad, but it is still a far-cry that the convincing thing: a demo, a quickstart tutorial, or a video.
Would love to try this, I mean, absolutely love!
But there's no mention of cost anywhere, and the link to the TOS is dead. So I'm not signing up just yet.
Just found the TOS while browsing.
The link under "sign up" (leading to https://flat.io/legal) was the one I referred to as broken.
(...) you grant us a worldwide, non-exclusive,
royalty-free and perpetual licence to use,
copy, reproduce, distribute, adapt, re-format,
modify, publish, translate, licence, sub-licence,
and exploit the User Content anywhere and in any
form for the purposes of providing our Service (...)
@gnud Thanks for the link, we will fix this shortly. This part of the ToS applies to the public & shared content that can be accessible by everyone on the internet ;-)
As written, it applies to the "music, sounds, text or information that you make available or create whilst using the Service" and gives Flat a "worldwide, nonexclusive,
royalty-free and perpetual licence to use, copy, reproduce, distribute, adapt, reformat, modify, publish, translate, licence, sub-licence, and exploit the User Content
anywhere and in any form." [1]
I can see why artists might be a little concerned about this.
We will update them in next few days / weeks to be more precise about private and non-shared content. This actually more apply to public content (e.g. you can see on "popular" pages).
At this stage, it's too early for this to be useful in my personal teaching practice. You undoubtedly have a feature comparison matrix. Hats off for what you've already accomplished.
AND I can see almost immediate use for your app at the stage it is in, with a slightly different angle!
My solfege colleagues would kill for a simple way to integrate music notation and text. Or a quick and easy way to build rich solfege exercises.
I suspect that that may be much easier both to pull off and monetise than yet another (and for now rather feature poor) music notation app.
Hey, check out my product Soundslice (soundslice.com), which I think might interest you. We're not a notation editor -- we let you combine (existing!) notation with audio/video and share that with students. See soundslice.com/teachers and let me know your thoughts, if you're up for it.
Github has special rendering for Markdown and CSV files. Would be awesome if they did the same for Lilypond and/or MusicXML and made such files playable.
Definitely, we already provide a transparent "fork" feature for teachers. With that they can easily distribute copies of their own documents to all their students.
Wow! It's Etherpad for music scores. Very impressive and very useful to people like me who don't have the interest in purchasing or maintaining something like Sibelius but want to explore musical ideas with other people. Hope it's not too expensive when it exits beta!
I'd like some way to write music like a piano roll or a pad system, or just some alternative to Western classical notation, which I find hard to translate to sounds in my head.
We use Node.js / Lua / Python, depending which services. For example for the engraving part we use some common code shared between the frontend/backend, same for the audio rendering.
Things is that Finale and Sibelius are expensive.
They're not very intuitive and convenient for beginners.
We do not expect to come as a pro software like them.
Finally we're only 1 year old they're more than 10. Give us some time ;)
I was confused about this, too. I added an "electric bass" part to the auto-created "my first score" and didn't see a "tab" option. Browsed through the "support -> editor" page and didn't see anything about tab parts in there, either. I only saw the tab option when creating a new piece.
Great to see the evolution to web. Struggle a bit with triples, specifically quarter note. The editor always wants to insert sixteenth note triplets.