[Colonel 32dll]

[Farottone]

Not worth it. You would end up with a ton of identical chords (when in fact they are different). See how autoreductions sometimes flatten chords between Hard and Medium and multiply the issue by 1,000. All of that after you’ve done days of research to convert guitar chords to 5-lane notes. However, you can prove me wrong: instead of programming anything, just come up with a table of conversion for all possible chords. I suspect you would quit after 5 minutes. ” src=”/wp-content/uploads/invision_emoticons/default_SA_wink.gif”>

[wingsoveramerica]

Is there any way that i can change the bit rate of OGG files when converting to Phase Shift?

[Dahnlee]

Not worth it. You would end up with a ton of identical chords (when in fact they are different). See how autoreductions sometimes flatten chords between Hard and Medium and multiply the issue by 1,000. All of that after you’ve done days of research to convert guitar chords to 5-lane notes. However, you can prove me wrong: instead of programming anything, just come up with a table of conversion for all possible chords. I suspect you would quit after 5 minutes. ” src=”/wp-content/uploads/invision_emoticons/default_SA_wink.gif” />

 

Okay, I do see your point about a chord conversion table not being the way to go, but that wasn’t exactly what I meant.

 

Maybe a better way to frame this idea is to look at it from the perspective of automating the MIDI process in the C3 tutorial (yes, I know you literally wrote the book on this) ” src=”/wp-content/uploads/invision_emoticons/default_SA_wink.gif” />

 

First, automating empty row removal would certainly be a timesaver.

 

Automating the process of “dumbing down” the chart is not as straightforward, but the two methods you point to in Section 9.3 could be implemented in a program. Even if you only program example A (the less subjective of the two), it still saves a lot of effort. Example B, by the way, is the point where I thought some of the RBN theory could be implemented.

 

The easiest approach would be to reduce all the guitar chords down to single “notes” using the mathematical mean of the MIDI note numbers (so that the trend of the pitch remains) and condense these notes down to five rows. Granted, it would then fall to the author to extrapolate chords by ear as needed.

 

Essentially, I’m suggesting something that would automatically convert a MIDI track into a rough chart that can be immediately pulled up in the RBN preview window and worked from there.

[Farottone]

The way I designed CAT is this: I used to do a process by hand, make a path to reproduce each step programmatically and then code in Python. If anything I was doing by hand wasn’t feasible programmatically I’d drop it. Reducing chords was one of the things I thought of, but too much of that process is the brain having to make unique decisions. Plus there is one big issue: MIDIs are never perfect, and if you go over a MIDI with defects it is essential that you do so with the original file, not with a dumbed down version that multiplies the severity of the inaccuracies.

 

In short, it would be great, but it doesn’t work, not even for bass, which rarely uses chords. ” src=”/wp-content/uploads/invision_emoticons/default_SA_wink.gif”>

[Dahnlee]

Sorry if I’m running this into the ground, but I’ve thought some more about the chord aspect. It would definitely be a mistake to go the route of an actual guitar chord equivalency table, but a program algorithm could characterize each composite midi note by:

 

1) mathematical mean (of midi components) = relative note placement

2) complexity (number of midi components) = number of buttons (one, two, or three)

3) expanse (delta between highest midi note and lowest midi note) = ex. GB chord (higher delta) vice RY chord (lower delta) given two buttons and the same relative note placement

 

Pretty sure these three characteristics would allow for relative RB chord mapping.

 

All that said, I was writing this while you were responding above. I recognize that you have put a lot more thought into this over time and I defer to your expertise

[TrojanNemo]

Thanks for the recent updates. There are a few requests I made, that I’m hoping will get added/fixed.

• Finding duplicates using the indexer. I appreciate it being added to setlist manager, but i’d like to know i have potential duplicates before they are in the game (ie reading cache file in setlist manager). This would be usable except for issue 2 on my list, which prevents me from dropping all my songs on setlist manager. Can the code be reused in the indexer ?
• Removal of the drag n drop “1 of 1” prompt for multiple files dropped on setlist manager.
• Batch search for songs with unpitched vocals.

quotes below, to refresh your memory, so you don’t have to search back through the post.

 

lastly, I’ve tried using event manager, i like it, but the one thing that is annoying is. After you enter a band name, select a song, then add it to the queue. The band name textbox is cleared. if someone wants to add multiple songs for “their band” they have to keep entering the band name for each song. it would be better if the band text box was a text drop down combobox, then you could quickly select your band, and only clear the song/artist fields when adding to the queue. Using a combobox for band also allows you to see which bands are in the list (for editing).

 

Those all sound reasonable and doable. Will add to the list.

 

Is there any way that i can change the bit rate of OGG files when converting to Phase Shift?

It was an option in the past, I don’t know why I ended up removing it, but it’s simply enough. Why are you not happy with the current quality? It’s higher quality (5) than HMX uses (3).

 

Sorry if I’m running this into the ground, but I’ve thought some more about the chord aspect. It would definitely be a mistake to go the route of an actual guitar chord equivalency table, but a program algorithm could characterize each composite midi note by:

 

1) mathematical mean (of midi components) = relative note placement

2) complexity (number of midi components) = number of buttons (one, two, or three)

3) expanse (delta between highest midi note and lowest midi note) = ex. GB chord (higher delta) vice RY chord (lower delta) given two buttons and the same relative note placement

 

Pretty sure these three characteristics would allow for relative RB chord mapping.

 

All that said, I was writing this while you were responding above. I recognize that you have put a lot more thought into this over time and I defer to your expertise

This is all a moot point as far me doing it, as I have zero experience with programming in for REAPER or with Python, so I would not be taking this on. Besides, farottone has infinitely more music theory knowledge and charting experience under his belt. If he doesn’t think it’s feasible, I’m fairly sure I wouldn’t be able to even wrap my head around the logic. Sorry, but this is not one I will be taking on.

[wingsoveramerica]

 

 

 

 

 

 

It was an option in the past, I don’t know why I ended up removing it, but it’s simply enough. Why are you not happy with the current quality? It’s higher quality (5) than HMX uses (3).

Whenever i compare the audio quality to what i hear in cPlayer, it doesn’t sound anywhere near as good. it’s like 128kbps mp3 compared to WAV.

[Farottone]

Sorry if I’m running this into the ground, but I’ve thought some more about the chord aspect. It would definitely be a mistake to go the route of an actual guitar chord equivalency table, but a program algorithm could characterize each composite midi note by:

 

1) mathematical mean (of midi components) = relative note placement

2) complexity (number of midi components) = number of buttons (one, two, or three)

3) expanse (delta between highest midi note and lowest midi note) = ex. GB chord (higher delta) vice RY chord (lower delta) given two buttons and the same relative note placement

 

Pretty sure these three characteristics would allow for relative RB chord mapping.

 

All that said, I was writing this while you were responding above. I recognize that you have put a lot more thought into this over time and I defer to your expertise

 

If you can write that algorithm and make it work, I’ll definitely look into integrating it. ” src=”/wp-content/uploads/invision_emoticons/default_SA_smile.gif”>

[TrojanNemo]

I’ve actually been wanting to pop the hood of CAT and see how it runs. I’m currently learning programming as part of my college degree including Python. Unfortunately I’m not terribly creative at coming up with helpful tools. The best I can think of is a preview for vocals but I couldn’t even begin to think of how to program that.

 

Someone tried that and it was a promising start, but then he gave up for some reason.

 

http://customscreators.com/index.php?/topic/12000-new-program-voxinator-a-rockband-vocals-visualizer

 

I would have loved him to continue. I took a look at his source code and he was definitely doing things in a much different way for me, but it worked. I took some of his ideas and implemented them in cPlayer, but his tool would have been much better for previewing vocals if he had stuck to it.

 

 

 

 

Whenever i compare the audio quality to what i hear in cPlayer, it doesn’t sound anywhere near as good. it’s like 128kbps mp3 compared to WAV.

 

That’s not right. Phase Shift converter uses OGG quality 5 when encoding, which is the equivalent of 192Kbps or so, since HMX uses quality 3 for Magma OGGs. cPlayer has access to the multitracks and to the tuning information from the DTA, so it should sound a bit closer to how it would be in game. Phase Shift converter does a decent job at trying to simulate that in the stereo file, but maybe the mix is a bit off? It definitely shouldn’t sound too different though. I have to look through the notes and comments as to why I removed it. Like I said, it’s very easy to implement a user option, but it was removed for a reason. I just don’t remember what that reason is.

[Bansheeflyer]

I would have loved him to continue. I took a look at his source code and he was definitely doing things in a much different way for me, but it worked. I took some of his ideas and implemented them in cPlayer, but his tool would have been much better for previewing vocals if he had stuck to it.

 

I might look at it myself. I doubt my skills are advanced enough to finish what he started but it wouldn’t hurt to see.

[wingsoveramerica]

That’s not right. Phase Shift converter uses OGG quality 5 when encoding, which is the equivalent of 192Kbps or so, since HMX uses quality 3 for Magma OGGs. cPlayer has access to the multitracks and to the tuning information from the DTA, so it should sound a bit closer to how it would be in game. Phase Shift converter does a decent job at trying to simulate that in the stereo file, but maybe the mix is a bit off? It definitely shouldn’t sound too different though. I have to look through the notes and comments as to why I removed it. Like I said, it’s very easy to implement a user option, but it was removed for a reason. I just don’t remember what that reason is.

 

For me it’s coming out varying from 112kbps to 155kbps.

[FujiSkunk]

Some observations that may or may not apply here.

 

Ogg Vorbis at Q5 strives for an average bitrate of 160kbps. Vorbis is a much better codec than MP3, so 160 for Vorbis is comparable to 192 for MP3, which is why the 192 number gets thrown around. Q3 strives for an average bitrate of 112, which in theory is about as good as MP3 at 128, and in my experience actually sounds much better.

 

It is not all that uncommon for Ogg Vorbis files to end up with an average bitrate noticeably higher or lower than 160kbps, even when encoded at Q5. I have one Q5-encoded song with a 200kbps average bit rate, and another with a 127kbps rate. Both stereo, both pulled from CD’s released in the late ’80s / early ’90s. It all depends on some very technical details about the original source audio that I’ve never bothered to learn for myself, but I appreciate that the codec doesn’t slave itself to a number and instead tries to deliver a consistent quality no matter how many bits are actually required, or not required.

 

And speaking of the original source audio, how your source audio is stored is a big factor in how the encoded audio will turn out. Basically, you’ll want to start with a source that has never been run through a “lossy” codec like Vorbis or MP3. Songs downloaded from iTunes, songs extracted from Flash or HTML5 videos, and most other music files you’ll get off the Internet have been run through a “lossy” codec. If you’re starting with compressed audio, you’re going to get lousy results when it gets compressed yet again. The few times I’ve tried to go the other way and create MP3’s out of Ogg Vorbis files, the results have been very “watery” despite the fact they were Q5-encoded Ogg files which to me were indistinguishable from the CD’s they were ripped from. When I made MP3’s directly from those CD’s instead, the results were much nicer to listen to.

[Busta Uppa]

Quick question – how do you change the in-game album art in an existing custom? I used to do it all the time, but C3 CON Tools is so full of features now that I’m a bit lost!

 

I’m thinking I need to use some combination of the Batch Extractor and CON Creator now, but I’m not sure.

[TrojanNemo]

You drop the CON file on the main screen of C3 CON Tools (or on the icon for C3 CON Tools) and it’ll open up. Then drag the replacement png_xbox file to the corresponding field. It’ll ask to overwrite, click yes. Then click Save. That should be it.

[Author]

Posts