The Amiga Future 143 was released on the March 5th.
HippoPlayer was just an example how it can work and on fast systems you don`t need that. The main idea (table: channel = n volume boost) you got already. So far we have three possible config settings: "autoboost calculation" (slow, but best result), "autoboost table" (not exact, guessing result but don`t need time) and "no autoboost" (current behaviour). Your idea "store user boost in the module and restore it when loading" is good but will only work modules you have saved with a comming DPB3 version.
In the past I`ve alredy mentioned OctamedSoundStudio`s "Sample List Editor" (not perfect but you will get the idea) and I think it is on the DBP feature list. Have a look at it and you know that it is the way to go. Even Protracker has a much better sample loading technic where you have a list of samples and just click on it and can played via keyboard. That the current DBP window close after each sample load is very uncomfortable and time consuming. About detailed features we can talk if you wish to change sample loading.
I think doing full autoboost at load is overkill. Then I've just implemented an option (switchable in program configuration) to set the master volume autoboost based on number of tracks. I've calculated exact autoboost for 20 or so modules and found that very simple formula:
estimated boost in dB = number of tracks / 2
works well. So, it will be in beta 27.
Finally we have some features listed ;-). The good news is "it is doable", the bad one "not trivial to implement", so I'd rather postpone it to some of 3.x updates.
It would be handy to know these features in advance, before redesign of this program part is done.
Hmm... I tested a 4 channel dbm with very loud samples. Autmoatic volume boost (Mixer) says 7,4 dB. With your formular it would be 2 dB. Strange to me.
Detailed sample load features: Take the OctamedSoundStudio as a base (is the source code available?) + Protracker "click on sample load it play it" what is nearly perfect. Before you start designing ask the DBP3 userbase (betatester?) for what they want. Would be better then asking me.
On the other hand module "Galaxy" by X-Ceed has 8 tracks, so my formula gives +4 dB, while the real autoboost is +2.5 dB. This only proves my point that calculating boost level from number of tracks with any magic formula is pure guessing.
DigiBooster Pro 2.x mixing engine relies on AHI for doing resampling, mixing and echo. DigiBooster 3 resamples, mixes and preforms echo inside its own engine, and sends ready stereo downmix stream to AHI. This makes a difference.
Hello dear DigiBooster3.0 - Programmers
Please provide at the final version an:
it should be an option in the settings of DB3.
e.g.: The user may should be able to choose at what time-periode (minutes) the program should automatic save the actual session.
that will prevent loss of very valuable time of new-programming patterns after amiga-crashes.
!!!Please take that Point SERIOUS!!!
As Pro-Audio-User i really really recommend to you to provide that option to DB3, because:
i often made the experience (since 1990) that the amiga may crash during slamming the keybord during a live-act, jam-session, or a pure production-session!!!!!
thats not good, because this limits the creative workflow.
it reeeeeealllyyyy dont makes fun to re-write a pattern during you are in "creative state", and i often loosed a "million dollar-pattern" because i dont remembered to save that masterpiece early enough...
(...you will know about what i am talking...)
so please insert that "auto-saving function"!
(i dont have any problems to become interrupted while the project is being saved! )
2. In this one i am also really interested:
I own "aegis sonix" - an amiga software from 1986.
this ancient old Amiga500-Programm allowes to make REALTIME effects like "cutoff" and "Resonance", besides much other standard-synthesizer-settings...wich sounds really good!!!
are you able to insert those two effects-routine-algorhythms in your programm, by using commandos like
"Q": for "Resonance", and:
"K" for "Cut-Off"???
those effects are essentially for all kind of electronic music...
those effects working in REALTIME on an Amiga500, with NOOO additional CPU-Power, besides the MC68000!!!!
So whats the problem to build that effects in your DB3???
You may have to decide if the auto-save function should overwrite the current file or create a new file with suffix extension (e.g. numbers). To be sure to save a go back possibility until DBP have an undo function.
Command K00 (Key off) is already used but filter effects would be nice to have. At least a plugin interface for developers so that other programmer can do it.
Good idea. I guess two options:
1. Time between autosaves.
2. How many versions of backups to keep.
will be OK.
Time of saving module is very short (For example I was working on 4 MB module on PowerBook yesterday, and saving the module takes about one second).
They are planned. Internally DSP effect mask of a track has 32 bits for a reason :-). However they probably won't make into 3.0, but some later 3.x update. I plan to add 3 filters:
1. Lowpass (which is the same as cut-off, I believe).
About parameters, I guess slope for lowpass and highpass may be fixed at 12 dB/octave, the frequency will be tunable realtime with effect parameter. "L" command may be used for lowpass, unfortunately "H" is taken already (yes, it is global volume slide), so some other letter have to be proposed. For now, if we leave small letters aside, following ones are free to use: I, J, L (it is envelope slide, but no plans to support it), M, N, O (risky, too similar to 0), Q, R, S (well, it is global pitch change, for now not implemented), T, U. Also there are 24 small letters waiting...
However resonance filter has more than one parameter, I guess, the most important one for realtime change is the frequency, but there is also gain, and Q factor. If I want to make all the three realtime tunable, I would need three effect letters. Alternatively gain and Q factor may be settable per-track, but not realtime (from DSP panel).
Yes, these filters are relatively cheap in terms of CPU load. Two-section lowpass or highpass filter needs two multiplications per audio sample.
None, except of time needed. I will try my best to get them into 3.0, but it may be they will be added later.
This will impose some serious limitation. For now, DigiBooster module is self-contained. It means that the module itself, plus a replayer is all you need to play the module. Once there is a way to add realtime plugins, this is over. It will be like Renoise then, where *.xrns file is not portable, as it depends on the set of VST plugins installed on a host. Then I personally do not consider plugins for player a great idea.
On the other hand plugins will be possible for the sample editor (both synthesizers and filters).