Hextracker (updated to 0.849B + BPM preview)

[tteclado]

[Regarding all the rest you refer:
- 50kHz player via the Yamaha / Printer port on STF: i do not know any case of that with 4 voices and volume control;
- (Disney Sound Source / Covox Speech Thing) are not known to me;

So if you have some infos regarding the items above, please post the related files or the links to them.

Covox Speech Thing - Wikipedia - http://en.wikipedia.org/wiki/Covox_Speech_Thing

I have been using the tracker AudioSculpture / Expose Software, using the output Yamaha / Printer, the program uses the output of the yamaha for channels 1 and 3, and can submit a DAC (Covox) channels 2 and 4, thus gives an interesting stereo ATARI ST.

[ljbk]

[Regarding all the rest you refer:
- 50kHz player via the Yamaha / Printer port on STF: i do not know any case of that with 4 voices and volume control;
- (Disney Sound Source / Covox Speech Thing) are not known to me;

So if you have some infos regarding the items above, please post the related files or the links to them.

Covox Speech Thing - Wikipedia - http://en.wikipedia.org/wiki/Covox_Speech_Thing

I have been using the tracker AudioSculpture / Expose Software, using the output Yamaha / Printer, the program uses the output of the yamaha for channels 1 and 3, and can submit a DAC (Covox) channels 2 and 4, thus gives an interesting stereo ATARI ST.

Hi !

Thanks for the reply.
Sorry for this late post. I have been quite busy lately.

Well, i have 2 questions.

1- Is Hextracker working via the Prosound output type with the Covox_Speech_Thing ?

2- How do you calibrate the volume between the YM ouput and the printer port device in case of stereo replay ? There is a high chance that you will get the volume on one pair of voices higher than the one for the other pair of voices, right ?

Bye,
Paulo.

[nativ]

Hi,

could a 4/6/8 or 16 channel output be made for Hextracker using a Covox (HexVox) ?

just a thought!

[ljbk]

Hi,

could a 4/6/8 or 16 channel output be made for Hextracker using a Covox (HexVox) ?

just a thought!

Hi !

If, as i understood, the Covox behaves as a cheap DAC attached to the Printer port then selecting the "Prosound" output type should work as in that case a 8 bit data is sent to the register 15 of the YM 2149. That 8 bit data is the result of the whole sound mixing and amplification.
No SW change should be required.

Now, to have a stereo output with YM on one side and Printer port on the other side or any other odd pairing like ST Replay and YM or ST Replay and Printer port, new output types must be allowed and specific code (rather small) as to be written for them.
The main problem for these cases is mainly a practical one: you have to ensure that for the same digital signal, you get the same, or at least a similar amplitude at the analogue output for both devices. For this, the ideal way would be to send a square wave to both devices and adjust the analogue signal amplitude for both devices. That is what i call calibration. But i think one needs an oscilloscope: a real one or a virtual one via a Line In port of a PC. A multimeter might do the job but not with a very good accuracy.

Bye,
Paulo.

[Hippy Dave]

... That is what i call calibration. But i think one needs an oscilloscope: a real one or a virtual one via a Line In port of a PC. A multimeter might do the job but not with a very good accuracy.

Bye,
Paulo.

A multimeter should work exceptionally well on the AC range with frequencies between 50 and 1KHz. Use the same square-wave frequency (eg. A-440 to E-660 Hz) and voltage range for your measurements.

[nativ]

Hi,

I actually ment would it be possible to build a covox type device and driver that support 4 or more individual outputs like an FA8?
could such drivers be compatibile with octalyser too?

[ljbk]

Hi,

I actually ment would it be possible to build a covox type device and driver that support 4 or more individual outputs like an FA8?
could such drivers be compatibile with octalyser too?

Hi !

I guess, you mean you would like to have multiple Covox type devices on the cartridge address space from $FA0000 to $FBFFFE, right ?
If that is the case, then you could build easyly a set of 7 Covox devices with A9 up to A15 selecting individually each device. With a bit more logic (gates) you can have much more devices.
In terms of software this is a much larger job as you will need a new path code for each number of combinations: 4, 6, 8, 12, 16 ... devices as you have to find out if you need to do any mixing at all at of which virtual voices or not.
The software would have to know how the set of devices would be mapped on the address space and the bit size (8 or more).

So if this is what you are thinking about, i would say that it is possible. But i can also say that it will not be implemented for Hextracker.
Sorry ...

Bye,
Paulo.

[yerzmyey]

Hi Paulo,
I've sent You a Private Message regarding the Hex.
Some strange things I encountered while Octalyser<=>HexTracker movements.

[ljbk]

Replay sent via PM some days ago.

Nothing strange with Hextracker behaviour.
As a general rule, if a MOD is not loaded there is a high chance that more memory is needed.

Bye,
Paulo.

[yerzmyey]

Hi,

just wanted to say I tested the newest version on original 520ST and it works perfectly well.
I don't know what's going on with STfm models, there must be some difference.
Still, now I have hardware read to work with the Hex.

Best wishes,
Y

[yerzmyey]

Actually You guyz could use the Hex and make something for this music contest:
http://atari-forum.com/viewtopic.php?f=19&t=21931
The best occasion for the Hex's "battlefield tests".

[ljbk]

Actually You guyz could use the Hex and make something for this music contest:
http://atari-forum.com/viewtopic.php?f=19&t=21931
The best occasion for the Hex's "battlefield tests".

Hi !

Congrats on your Funky ST MOD !
Optimal V/Q on Hextracker calculated by the tool is 5.3 in Mono (as you prefer) to have less than 1% distortion.

Sorry not to have updated/improved the player/editor lately but i have been too busy with other stuff.

Cheers,
Paulo.

[yerzmyey]

ljbk:

>> The best occasion for the Hex's "battlefield tests". [/quote]
---------------
> Hi !
> Congrats on your Funky ST MOD !
-----------
Heh, thanx. Yes, there's my Hex/Octa song in the thread http://atari-forum.com/viewtopic.php?f=19&t=21931 indeed.
It is not only a very powerful (most?) engine, but also I have to admit I'm pretty satisfied with the outcome. Great tool, man. And btw, the song is recorded form STe, however it sounds really clear and good from regular ST!

> Optimal V/Q on Hextracker calculated by the tool is 5.3 in Mono (as you prefer) to have less than 1% distortion.
Sorry not to have updated/improved the player/editor lately but i have been too busy with other stuff.
----------------
Oh! There is any util that calculates the best settings??? (Desirably for YM )??? It would be a great thing!!!

> Cheers,
Paulo.
-----------
Once again manmy thanks for Your work - it allows to produce incredible things on ST!
Greetz,
Y

[ljbk]

Hi !

To calculate the best VQ ouput just use the DIST option.
There are only 2 optimal values: one for Mono and one for Stereo.

Using Steem:
----------------
Set the virtual machine to 40 MHz and select in Hextracker 8 bit, DIST and the maximum value for F.
Now play the complete MOD from start until it loops.
When playing stops, a value will appear next to DIST on the right where there is normally 0.0 .

Using the real machine:
-----------------------------
Set your normal playing conditions.
Reduce the replay frequency by about 15%(Mono) or 25%(Stereo) (Check /*Just Dist */ in the manual).
Now play the complete MOD from start until it loops.
When playing stops, a value will appear next to DIST on the right where there is normally 0.0 .


Example with your MOD:
Normally we manage 18 KHz with all option off (17.5 KHz on slower STs).
Now i select F = 14985 (i could have used 15360) and DIST.
I play and wait 2:57.
When i stop playing the value 5.3 appears.
The value will be the same for all Mono outputs and it varies very little with the replay frequency and 7bit/8bit option.

Now for STE Stereo:
F= 16.7 KHz.
I play and wait 2:57.
When i stop playing the value 3.6 appears.
The value will be the same for all Stereo outputs and it varies very little with the replay frequency and 7bit/8bit option.

So for your MOD on a 1040 STE using Hextracker you should use:
- STE Mono 25 KHz and V/Q = 5.3;
- STE Stereo 25 KHz and V/Q = 3.6;
- YM (Mono) 18(or 17.5) KHz and V/Q = 5.3;
You can also use real 8 bit mixing (selecting 8 bit) instead of 7 bit mixing (The Player uses 6 bit mixing with 8 voices to achieve the 50 KHz replay) but then you have to reduce the replay frequency (STE 16.7 KHz / YM 14.2(or 13.9) KHz).

Just select the config / player that sounds better to you .


Enjoy,
Paulo.

Edit: copy/paste error (Mono => Stereo)

[yerzmyey]

Aha. So the option is simply implemented in Hex-Tracker v0.843.
I will check out the option soon!

[ljbk]

Aha. So the option is simply implemented in Hex-Tracker v0.843.
I will check out the option soon!

Yes it is.
This specific feature together with the V/Q factor are exclusive to Hextracker as far as i know and it was one of the specific things i shown to the guys in Uppsala in 1992 as i describe in my first post in this thread.
When saving to the Hextracker format, the last used value is stored in the file with a bit telling if that value was for Mono or Stereo playing.
When loading the file in case of Mono value for Stereo Output Type and vice versa a 1.5 factor is used to convert the file value. The Stereo VQ is ALWAYS smaller than the Mono VQ for the same MOD of course.

In the early 90's it was common to have a 2.0 VQ hardcoded in the MOD players (for 4 voices).
As you are mixing 4 values from 0 to 255, so you can get values from 0 to 1020. With a 2.0 VQ hardcoded, values below 256 were considered as to be 256 and values above 767 were considered to be as 767 and then you would apply your YM table to the 512 values from 256 to 767.
This worked correctly for most of the MODs but then later some problems appeared like:
1- Stereo meant 2 mixes of 2 voices and then an hardcoded VQ of 2.0 lead to much distorting while no amplification meant low volume level;
2- Increase use of volume control ($A and $C commands) lead to a lower global volume level for the MOD and then a higher amplification was needed but an hardcoded VQ of 4.0 (considering only values from 384 to 639) was leading to much distorting;
3- For modules where samples were already amplified with some distorting like heavy metal guitars, basses and so on, a VQ factor of 2.0 is already too much;
The variable V/Q feature and the DIST function to calculate it, solved all these problems plus one that was exclusive to Hextracker: the variable number of voices to mix.

Taking your MOD as example, a VQ of 5.3 (in hex) = 5 + (3/16) = 5.1875 means that the mixing is amplified more than 5 times in Mono.
In case of Stereo as we are mixing 2 times 4 voices, the VQ is lower at 3.6 (in hex) = 3 + (6/16) = 3.375 . The calculated value shown is the lower of the 2 values calculated for left speaker and right speaker.
As you can see, the VQ for Stereo is smaller and the ratio between the two is: 5.1875 / 3.375 = 1.537.
That is very close to the assumed 1.5 value refered above ...



Cheers,
Paulo.

[ljbk]

Hi !

To give you another example, i just converted the speed030 module by 505 to MOD format using OpenMPT 1.19.02.00 .
Testing the resulting MOD on Hextracker (loosing at least the stereo panning (command $8)), one gets:

Mono: VQ=3.6
Stereo: VQ=2.4

So the configs should be
YM: 17.5 KHz VQ=3.6
STE Mono: 25 KHz VQ=3.6
STE Stereo: 25 KHz VQ=2.4

Again 3.6(hex) / 2.4(hex) = 1.5 (decimal) !

Nice MOD BTW.

Bye,
Paulo.

[ljbk]

Hi !

After a very long pause probably followed by another one, here's a 47.2 KHz 100%CPU replay for 4 voices.
So it is an YM50K player but at 47.2 KHz for the "impossible" modules at 50 KHz.
You can now listen to the Agony Amiga game intro music on your 1MB STF for example: i like very much this piano mod.
In the zip, you can find the original agonyint.mod and the Hextracker mod agonyh.mod where VQ is set to the optimum value 3.7 .

Enjoy it.

Cheers,
Paulo.

[Hippy Dave]

Hi !

After a very long pause probably followed by another one, here's a 47.2 KHz 100%CPU replay for 4 voices.
So it is an YM50K player but at 47.2 KHz for the "impossible" modules at 50 KHz.
You can now listen to the Agony Amiga game intro music on your 1MB STF for example: i like very much this piano mod.
In the zip, you can find the original agonyint.mod and the Hextracker mod agonyh.mod where VQ is set to the optimum value 3.7 .

Enjoy it.

Cheers,
Paulo.

I enjoyed the excellent music and have been listening to chip-tunes
a lot recently because there is hidden gold there. I've also been playing
my guitars more too.

Quartet Mono is sounds a bit noisier than ST Replay.
What is the difference?

How do I exit? ...reset?

[ljbk]

I enjoyed the excellent music and have been listening to chip-tunes
a lot recently because there is hidden gold there. I've also been playing
my guitars more too.

Quartet Mono is sounds a bit noisier than ST Replay.
What is the difference?

How do I exit? ...reset?

Hi !

Exit is by done left clicking on the top left corner like in Hextracker.
The diference between Quartet replay and ST Replay is the difference between two sound tables and two methods: YM voices ON for Quartet (like the program Quartet does) and YM voices OFF for ST Replay. I remember we debated that a lot in the Hatari thread.

Cheers,
Paulo.

[Hippy Dave]

I enjoyed the excellent music and have been listening to chip-tunes
a lot recently because there is hidden gold there. I've also been playing
my guitars more too.

Quartet Mono is sounds a bit noisier than ST Replay.
What is the difference?

How do I exit? ...reset?

Hi !

Exit is by done left clicking on the top left corner like in Hextracker.
The diference between Quartet replay and ST Replay is the difference between two sound tables and two methods: YM voices ON for Quartet (like the program Quartet does) and YM voices OFF for ST Replay. I remember we debated that a lot in the Hatari thread.

Cheers,
Paulo.

Thanks. By-the-way, I used Hextracker Hatari in ST mode and all went well.
I remember the debate and didn't understand because I'm not a YM2149 tracker expert.

Are Voices On = 125KHz AC; and Voices Off = 0Hz DC ?

If this is the case, I understand exactly what you are doing. Hatari currently
demodulates everything above 80% nyqist rate, and returns correct values for
Voices Off.

[FedePede04]

Hi !

After a very long pause probably followed by another one, here's a 47.2 KHz 100%CPU replay for 4 voices.
So it is an YM50K player but at 47.2 KHz for the "impossible" modules at 50 KHz.
You can now listen to the Agony Amiga game intro music on your 1MB STF for example: i like very much this piano mod.
In the zip, you can find the original agonyint.mod and the Hextracker mod agonyh.mod where VQ is set to the optimum value 3.7 .

Enjoy it.

Cheers,
Paulo.

Downloading...Downloading...Downloading...Downloading...Downloading...Downloading...Downloading...Downloading...Downloading

i just found my old 56k modem (no just kidding)
Thanks Paulo, good to see a new version.

Cheers
Peter

[Scarlettkitten]

Thanks for the update, I look forward to trying it out.

[yerzmyey]

Thx a lot dude. I will check it out too.

[Hippy Dave]

I enjoyed the excellent music and have been listening to chip-tunes
a lot recently because there is hidden gold there. I've also been playing
my guitars more too.

Quartet Mono is sounds a bit noisier than ST Replay.
What is the difference?

How do I exit? ...reset?

Hi !

Exit is by done left clicking on the top left corner like in Hextracker.
The diference between Quartet replay and ST Replay is the difference between two sound tables and two methods: YM voices ON for Quartet (like the program Quartet does) and YM voices OFF for ST Replay. I remember we debated that a lot in the Hatari thread.

Cheers,
Paulo.

Thanks. By-the-way, I used Hextracker Hatari in ST mode and all went well.
I remember the debate and didn't understand because I'm not a YM2149 tracker expert.

Are Voices On = 125KHz AC; and Voices Off = 0Hz DC ?

If this is the case, I understand exactly what you are doing. Hatari currently
demodulates everything above 80% nyqist rate, and returns correct values for
Voices Off.

Hi Paulo,
I read your informative explanations from page 2 onwards of the thread:
"Hatari version 1.5.0 has been released". My questions were answered.
Your explanations and measurements really help in emulating high quality
YM2149 sound.
David