Recent Posts

Pages: 1 2 [3] 4 5 ... 10
21
Development Corner / Re: OpenMPT 1.26 releases
« Last post by Saga Musix on August 14, 2017, 16:55:57 »
Another small bugfix has been released to address one annoying bug introduced in the previous version: Updating the octave of a note below middle-C no longer worked as intended.
22
Free Music Downloads / Re: [Synthpop / Dance] Disassembly (it/ogg)
« Last post by Buckmaster on August 14, 2017, 07:24:00 »
Nice track Saga, thanks for sharing man!
23
General Chatter / Re: Open-source and free software
« Last post by arseniiv on August 14, 2017, 05:38:55 »
Thanks. However, when selecting pure MIDI, latency is perfectly small, and it’s just what I need. :)
24
General Chatter / Re: Open-source and free software
« Last post by monsterovich on August 13, 2017, 22:57:21 »
Is it possibly my Wi-Fi?

There is always a possibility. To decrease the latency I even tried to connect my smartphone to PC directly through the wi-fi-spot and had a good success (use this scheme: laptop - host, smartphone - client, otherwise you will experience major lags due to bad hardware on your smartphone). There is another way - just use the lan cable for your PC, it always works without issues.
25
General Chatter / Re: Open-source and free software
« Last post by arseniiv on August 13, 2017, 15:25:41 »
I realise it’s off topic here, but just can’t add one more postscript:

P.⁴ S. As I deduce, there’s currently no way to change preset on client startup. It’s a bit frustrating that changing preset needs editing the setting file manually. Is startup selection an option for newer versions?
27
General Chatter / Re: Open-source and free software
« Last post by arseniiv on August 12, 2017, 22:31:49 »
@monsterovich It seems you’ve added a MIDI port with some magic there. Do I need to use some other software to do it, and how? Your app deserves more attention from the community (and more documentation as well ;)) I’ve uncommented the lines in launchpadd.ini and replaced the setting with miditest, but at the time I have default Windows GM synth deactivated, so I couldn’t test if the server receives anything, but I believe so, and anyway it’s cool.

P. S. Oh, I’ve been wrong about the docs, now I found the wiki. Will read.

P. P. S. Now at least I’ve configured client & server to play sounds at their own without MIDI. :) Cool indeed, but unfortunately latency is too high. Is it possibly my Wi-Fi?

P.³ S. Yay, just some loopMIDI and it works like a charm! Totally inaudible latency when receiving in Tracktion! It seems I’m going to have to learn some MIDI message codes… :D
28
Development Corner / Re: OpenMPT 1.26 releases
« Last post by Saga Musix on August 12, 2017, 15:22:09 »
Yet more bugfixes! A couple of possible crashes and other problems have been identified in the last few weeks and the bugfixes have been backported to OpenMPT 1.26.13.00. Other small improvements have been added as well.
29
Cleanup -> Instruments -> Remove all (convert to samples) does exactly that. It re-maps the instrument numbers in the patterns, but I should mention that envelopes are not translated to pattern commands. Any instrument properties apart from the note->sample and note->note mapping are lost.

So that's how that could be done, thank you!

I should start by saying that the "transpose" field has no specific meaning in the S3M/IT/MPTM format; it's only there because in the MOD and XM format, samples have a transpose and finetune value, so the space in the GUI is already there. The Transpose field can be used in S3M/IT/MPTM to quickly change the sampling frequency in semitones but the content of the field is not really relevant.

Now for the origin: In the MOD format, there was no way to change the sample transposition. The middle-c sampling frequency was always 8272 Hz +/- one semitone.
Fasttracker 2 (XM) made this a bit more flexible, by specifying a transposition offset from middle-C (8363 Hz in XM, as it uses Amiga NTSC frequencies rather than PAL frequencies). So basically the way the transposition is displayed in OpenMPT means "playing a C-5 in the pattern really plays a [transpose value]". So if the transpose value is C-5, C-5 is played as C-5 (8363 Hz). If thranspose is C-6, then C-5 is played as C-6 (16727 Hz).

Back to the S3M/IT/MPTM format, the transpose value basically works the same there - if your sampling frequency is 16727 Hz, then transposition in C-6, if your sampling frequency is 44100 Hz, the transposition value is E-7, etc.

I hope this makes sense somehow. There are plans to modify this system a bit (by being able to specify the root note of a sample), but they will most likely not make it into OpenMPT 1.27.

Now I perfectly understand why the algorithm I described earlier could be used to find out the Transpose note based on sample rate. On Amiga the C5 note was sampled with sample rate of 8363 Hz meaning the bandwidth that could be recorded was up to 4181.5 Hz which is 8x523.3 Hz (8xfrequency of the C5 sine wave). 8 times higher bandwitdth than the fundamental frequency of the note allowed to record the waveform up to the 8th harmonic - which can reproduce the timbre quite well. Since 8 = 2^3, that 'three' explains the 'three' in the algorithm I've written in the last post: 'The note displayed as the Transpose value is the note which corresponds to the highest frequency that could be recorded using the given Frequency (sample rate) - transposed three octaves down.'.

I couldn't decipher the meaning of the Transpose because the notes are usually put in the cotext with the frequency of the recorded signal and not in the context with the sampling rate. When I saw Sampling rate = x Hz I was automatically reasoning: 'The highest frequency of the tone is x/2 Hz'. But everything is clear now that you explained - thank you very much for such a detailed explanation.

Yes, specifying the root note of the sample would be very straightforward way to use the imported samples. It would be something as being able to 'say': 'I have a sample of note x of some voice/instrument', and the sample would then be tuned to sound right. Since there already is a FFT analysing routine within the Sample Tuner, the root note field could be pre-filled with the predetermined value which could be accepted or changed by the user. However, the frequency field should then refer to the detected frequency of the root note of the sample rather than to the sample rate. That way if the user would, for example, import the sample containing the tone with fundamental frequency 440 Hz and FFT algorithm would detect e.g. 452 Hz - the user could 'say': 'No, that is not 452 Hz, I know for sure that is 440 Hz'.
30
Speaking of Instruments, I wanted to try ADSR envelopes so the Instrument mode way activated. Later I wanted to revert back to the Sample mode but couldn't find a way to do that. Ifter deleting the Instruments I tried everything to reenable using only the samples but didn't find a solution and by searching the docs and the forum I found only how to enable Instrument mode - but nothing about going the other way.
Cleanup -> Instruments -> Remove all (convert to samples) does exactly that. It re-maps the instrument numbers in the patterns, but I should mention that envelopes are not translated to pattern commands. Any instrument properties apart from the note->sample and note->note mapping are lost.

Let us see if I really understand how that works:

If I understand correctly, OpenMPT expects the imported sample of musical instrumet to be C5. If  I don't have a sample tuned to C5, I can change the pitch of the sample before importing it (to be C5) and it would sound right. If the sample is not tuned before importing I can tune it by changing the sample rate (the frequency) or by using Sample Tuner, or by using Pitch Shift (if I don't want to change the sample rate). So the best would be if I can sample the C5 note of the instrument using for example 44.1 kHz sample rate and no additional tuning would be required.

Next, for the Frequency and the Transpose fields - as I understand now upon reading your answer - the Frequency is in fact sample rate and the Transpose value (note) is calculated based on sample rate value. So if I imported the sample of A4 note I must shift the sample rate 3 half-tones up so the pitch would be C5 - as expected by OpenMPT. If the sample rate was 44100 Hz the Transpose value would be E7, meaning I have to choose G7 which is 3 half-steps up from the E7.
So far this is all correct.

Now, what I would like to know is how exactly is E7 calculated starting from the 44100 Hz as the input.
I should start by saying that the "transpose" field has no specific meaning in the S3M/IT/MPTM format; it's only there because in the MOD and XM format, samples have a transpose and finetune value, so the space in the GUI is already there. The Transpose field can be used in S3M/IT/MPTM to quickly change the sampling frequency in semitones but the content of the field is not really relevant.

Now for the origin: In the MOD format, there was no way to change the sample transposition. The middle-c sampling frequency was always 8272 Hz +/- one semitone.
Fasttracker 2 (XM) made this a bit more flexible, by specifying a transposition offset from middle-C (8363 Hz in XM, as it uses Amiga NTSC frequencies rather than PAL frequencies). So basically the way the transposition is displayed in OpenMPT means "playing a C-5 in the pattern really plays a [transpose value]". So if the transpose value is C-5, C-5 is played as C-5 (8363 Hz). If thranspose is C-6, then C-5 is played as C-6 (16727 Hz).

Back to the S3M/IT/MPTM format, the transpose value basically works the same there - if your sampling frequency is 16727 Hz, then transposition in C-6, if your sampling frequency is 44100 Hz, the transposition value is E-7, etc.

I hope this makes sense somehow. There are plans to modify this system a bit (by being able to specify the root note of a sample), but they will most likely not make it into OpenMPT 1.27.
Pages: 1 2 [3] 4 5 ... 10