OctaMED Sound Studio Install Issues & Advice.

[Karlos]

I used to use OctaMED Sound Studio all the time back in the latter half of the 90's. Since then I moved on to PC where I basically got hooked because of a handful of software that just does not exist on ANY OTHER Operating System and I can not live without them.

While having this extra power I began using Open Mod Plug Tracker and my OctaMED use fell to the way side, but...

...here we are, years later and I'm beginning to tinker with the old hardware again and I was that particular sound that only a real Amiga can produce in that unique Amiga flavor.

So I'm installing OctaMED S.S., but on disc two, it turns out that the Installer gives me an error.

It looks like the installer failed to either create or locate a drawer or assign. Have a look at the line of the script to see if you can fix it.

I think I installed mine manually. In the end, it only needs an assign or two anyway IIRC.

In regards to advice, what settings would you veteran users suggest for an Amiga 1200 030/40Mhz with 16Mb of RAM.

On an 020 with 4MB of fast ram I had no problems with 8-channel mixed mode at 27000Hz (can't remember the exact value, it is the upper limit of Paula for PAL). On my 68040, I've played over 32 channels in mixed mode before, although nowadays most of my stuff uses MIDI.

And what setting would you suggest for an Amiga 600 68000 with 4Mb of RAM total?

By this I mean OctaMED's Frequency settings, channels, halfing, sample quality, size, and all that jazz.

For a 68000, you are probably best sticking to regular 4 channel mode or using it as a MIDI sequencer. You will probably get away with turning on FastMemPlay though, increasing the total sample size you can get away with.

BTW, what is Sample halfing anyhow, and should I use it, because it only seems to make the samples sound like crap.

Sample halving is used in legacy 5-8 channel mode. In this mode, 2 "software" channels are mixed into 1 hardware 8-bit channel, requiring that the overall sample output never exceeds the 8-bit range. The only way to guarantee that is to half the volume of the samples used on those channels. This isn't needed in mix mode, which uses a full software mixer for all channels.

If you want my advice, don't use the old 5-8 channel mode. Use the mixed mode on your 030. It's almost certainly faster and has far fewer limitations.

Finally, why is it that when I am using OctaMED Sound Studio, that it sometimes begin to put out garbage instead of the sounds it is supposed to? The only way I've found to resolve this is to reboot.

Again, thankx in advance!

That sounds like a bug. I've only had that happen in non-Mixed 5-8 channel mode with very short sample loops, but only rarely.

[Karlos]

The version I'm using is the one that was given away by CU Amiga magazine (1.03c, I think). I recall trying a beta v2, but it was far from complete.

[Karlos]

Surely the popup isn't that annoying?

Anyway, regarding OS4, it runs just fine - at least as far as sample playback goes. MIDI, I haven't got working yet (probably because OctaMED SS bangs CIA/serial port directly for MIDI, I guess). In another thread, I think you were asking me about the instrument properties. The screenshot I added was running on OS4.1.

The way to get it working on OS4 is to use a tool called "The Maestrix" (or some such pun on the matrix) which emulates one of the supported soundcards on AHI. This application is also 68K and therefore works with classics/UAE as well (confirmed) and I expect it works in MorphOS too.

[Karlos]

Regarding the noise issue, there's nothing wrong with the IFF format itself. Obviously 8-bit sound has a rather limited resolution and you will notice quantisation noise with any 8-bit linear sound format (note that there are 8-bit encodings that are logarithmically delta encoded, which provide better SNR (signal-to-noise ratio) but aren't really supported as general sample formats).

The chances are that many .iff sound files you come across have been captured on an old 8-bit parallel port sound digitizer (or so we used to call them in the good old days). They didn't generally have the best SNR properties to begin with and the noise level will be in the least significant bit(s) of the sample. Add to that the fact that a lot of those sounds were captured at, or resampled to the old pro-tracker C-3 note playback rate (which is about 17kHz IIRC), then you get aliasing noise too. End result are audible hiss, ringing and other artifacts.

If you've used a 16-bit .wav or .aiff sound source, the chances are it's been captured on a reasonable recording device, perhaps even professionally, at a CD quality or better sampling rate. Scaling down to 8-bit adds quantisation noise again, but the original signal noise, which should have been in the least significant bit(s) of the 16-bit data, is now off the small end of the scale. That is to say, none of the noise in the original 16-bit recording, assuming it was recorded properly, can realistically be present in your 8-bit version as it doesn't have the precision to represent it. Only quantisation noise should be present. If you downsample the rate properly (using antialiasing) from 44kHz to say 22kHz - conveniently close to the F-3 note, you shouldn't get too much audible aliasing noise either.

Note that you can use 44kHz samples anyway, if you intend to use mix mode. It's nice  if you want to do a direct-to-disk version of a track, since the extra quality is noticable in the end result.

In short, an 8-bit sample downsampled from a 44kHz 16-bit source will always sound cleaner than one sampled on an 8-bit parallel port

[Karlos]

If you read up on sampling theory, it'll make more sense. The salient points are:

1) The sample rate should be at least 2x higher than the highest frequency component of the sound being sampled. So, don't use low sample rates for bright sounds (cymbals and the like). Keep them 22kHz or better if possible.

2) A 16-bit sampler basically has 256x the precision, per sample, than an 8-bit one. Assuming the background noise level only affects the least significant bit(s) of the sample, conversion from 16-bit to 8-bit produces a better quality than sampling at 8-bit. However:

3) Reducing the bit depth of a sample introduces quantisation noise.

If you are used to 8-bit sample quality, you won't really notice the effects of (3) so much but you should at least be able to tell the difference from an 8-bit sample that was sampled at 8-bit and an 8-bit sample that was reduced from 16-bit.

[Karlos]

@XDelusion

In order to play 16-bit samples in OctamedSS, you'll need to enable mix mode (or alternatively have one of the sound cards for which there is built-in support. See the earlier post about using The Maestrix for 16-bit AHI output). Otherwise, the 16-bit sample data is misinterpreted as 8-bit. The first (most significant) byte of the 16-bit word in each sample probably plays fine. It's the second byte (least significant) that will be misinterpreted as another 8-bit sample, which will almost certainly have no sensible relationship to the value it just played, producing some ungodly screech.

Unfortunately, mix mode on vanilla 68000 probably isn't going to work well as it needs too much CPU.

[Karlos]

I wish there was some library of good 8bit sounds on the web, you know of anything like that Karlos?

Well I might have to get this...

http://www.retrothing.com/2010/02/8-bit-weapon-chiptune-sound-library.html

There used to be a collection of SoundTracker sample disks back in the day. You could tell when they'd been used in .mod files, they were usually called ST-.

In the end, I've gathered my own sample library over the years. More often than not, these days, I rely on MIDI stuff controlled by octamed rather than it's own sample playback.

These days, you can get high sample libraries on CD in .wav / .aiff format.


Speaking of MIDI, there's a nice MIDI triggered sample playback tool called MIDI-In which uses AHI for output. What I've done in the past is to use a spare A1200 (with just fast ram) as the sequencer, running OctaMED SS to control various other midi gear, including my main A1200T, which has 256MB of RAM running MIDI In. The sound output in calibrated 14-bit mode is superior to what OctaMED SS itself can produce in realtime and furthermore, MIDI In is a bit more capable in that it can replay 16-bit samples without some of the OctaMED limitations. The sound output of that was routed through my MIDI tone gen which has AD inputs (as well as providing the remainder of the MIDI sound). A bit fiddly but the end result was quite nice

[Karlos]

The first step is to set 1-64Ch Mix Mode. You will find this in the Song options window (see attached image) which you get via the Song -> Set Options menu.

In the window you should see that you have the option of 4Ch/MIDI, 5/6/7/8Ch (these are the old OctaMED style channels, where 2 software channels are mixed into one 8-bit hardware channel) and finally 1-64Ch Mix Mode. The latter also supports MIDI too

The "high quality" mode only applies to the old 5-8Ch mode. It has no effect on mix mode.

[Karlos]

The second step is to configure your preferred mixing setup. Choose the Settings -> Mixing menu item. You should get the window shown in the attachment.

In this window, you need to choose your output mode, mixing rate, number of channels, panning etc.

Obviously, the more channels you choose, the more CPU is required. I had no problems with 8 channel 14-bit stereo at 28kHz (highest you can get in PAL mode) on the basic A1200 with added Fast RAM. Your 030 should be able to handle it no problem. I've had 32 channels on the 040 without any real issues.

The smoothing option uses interpolation when playing back samples at a rate lower than the mixing frequency. This produces a cleaner sound but the CPU load goes up dramatically with the number of channels. Unless you have a very fast 68060, forget it for realtime playback, use it when you want to try the Disk 16-bit mode.

As you increase the number of channels, OctaMED SS recomputes it's internal mixing tables to make sure that you don't clip the output. This causes it to get progressively quieter. You can compensate for this using the volume adjust slider, which boosts the output beyond what is theoretically the safe clipping limit. In practise, it's almost always a good idea to do so. In 16-channel mode, I've used volume adjusts of 350% without clipping.

The mix buffer size, is as the name suggests, size of the mixing buffer. Larger mix buffers decrease stutter when CPU load is high, but also increase latency and decrease responsiveness. The only time I've used higher mix buffers is when using direct-to-disk output.

If you click the panning button, you'll get a sub window with sliders for each of the channels. Panning is from -15 (extreme left) to +15 (extreme right). The "free panning" option limits the volume to avoid clipping issues again. I almost always have it turned off and rely on the overall volume adjust to get the maximum non-clipped playback volume.

The effects button opens another window with a few basic SFX that apply globally, such as stereo widening and echo. To be honest, I've not much use for those, since I route my audio through external gear anyway.

Hopefully this answers a few questions

[Karlos]

One final point to note. When you enable Mix mode, the note range for sample playback is increased by 2 octaves in the low register. That is to say, the pitch you'd get from C-3 in 4 channel mode is now 2 octaves lower. To get the same pitch, you need to use note C-5.

You'll encounter this problem if you load an old 4 channel mod and enable mix mode and try to play it. To correct it, you need to use the Edit -> Transpose option which opens the transpose window. You'd need to globally increase all notes by 2 octaves.

[Karlos]

Also if you could ever put up a youtube video of your midi set up in action, that would be awesome!

I've only really got my digital camera and if old experiments are anything to go by, the quality isn't great.

I'm still trying to wrap my mind around midi on top of OctaMED and eventually want to control my 2600 through my Amiga and perhaps set things up sort of like how you've got it. From the sounds of it, that's the way to go, but a video would help me visualize, since I'm not totally sure I fully understand how your set up works.

I've made a quick diagram of my two most often used configurations:



This is my usual day-to-day configuration. For the most part, I rely on MIDI to control the tone generator, with just a few samples perhaps from OctaMED itself. My tone generator has two MIDI inputs as it supports 32 parts. It allows you to map any number of parts to any of the 16 channels from either input. Live input from the controller keyboard is routed out of the tone generator and back into OctaMED so I can capture live performance data. Once a composition is complete, I then connect the MIDI Out of the Tone Generator instead of the MIDI Thru and dump all the SysEx data to OctaMED so it can be saved off as part of the current song.

Here (mp3 2.7MB) is an example track that uses only MIDI (no samples) that was created in this configuration. It's a remake of the Oscar theme tune (it was the first AGA game I played on my A1200, so I was feeling nostalgic).

Here (mp3 9.5MB) is an example track that uses MIDI and regular Paula 4-channel stereo with 8-bit samples only. The 4-channel raw output from Paula is processed by the Tone Gen to adjust panning, add reverb and so on.



The above config is something I experimented with when I wanted to try some really sample-heavy stuff. For this, I used my rather less powerful regular A1200, with just a 4MB memory expansion in the trapdoor. In this configuration, OctaMED is not playing any samples itself. It is acting as a pure MIDI sequencer. One of the MIDI outputs is sent to the Tone Generator, as before, and the other is sent to the MIDI Input of the A1200T. The A1200T is, in this configuration, running a piece of software called (confusingly) "MIDI In". This piece of software is actually rather good and is best described  as a "software sampler". It handles many samples mapped to arbitrary MIDI channels with full support for keyboard split multisamples (ie you might have a drumkit with a different sample on every note). It accepts all the usual MIDI performance stuff: note velocity, modulation, pitchbend etc., making it very easy to control from a keyboard or from OctaMED. Lastly, it supports both 8/16-bit samples without any of OctaMED's own internal 16-bit replay limitations as it uses AHI for playback. With calibrated 14-bit output, it can sound surprisingly good.

Here (mp3 3.6MB) is an example track created on the above configuration. It uses several large MIDI triggered 16-bit samples being played by the "MIDI In" application on my A1200T. AHI calibrated 14-bit output was used and processed through the Tone Generator to apply volume compression (was designed to sound deliberately overcompressed to make the percussion punchy). All of the sequencing and effect controls were handled by the basic A1200.

In both configurations, the audio output from the A1200T is connected to the A/D input of my Tone Generator, which not only results in the sound being mixed, but also allows me to add realtime effects to the incoming Amiga audio stream which in turn can be controlled via MIDI too.

[Karlos]

^ And yes, to get that working I have 2 MIDI interfaces, one for each machine. Both have 1x MIDI In, 1x MIDI Thru, 3x MIDI Out (sadly, not separately addressable).

If you just want to attach 2 Amiga machines together and use them as a MIDI sequencer / MIDI device pair, a plain serial cable will do the trick. Traditional MIDI is basically RS232 sent down a 5-pin DIN lead anyway.

[Karlos]

In regards to Mix Mode on OctaMED...

...that's what I had done, but I still get noise, so I'm assuming an 020 with minimal RAM is not suitable for the job. Hence the reason I'm going to wait for my 030 to arrive before I mess with higher quality samples at all, in the mean time, I'm going to try to use that static constructively.

Dunno about that. I have definitely used it on an 020, even without any additional fast ram. It isn't very useful in that case, but you can certainly use 4 channels in mix mode, even if the only reason you use it is to adjust their stereo position.

With fast ram, on an 020, I've used mix mode with up to 12 channels before at 14-bit/22kHz. I used to use that configuration. One of the reasons I bought my first accelerator card was so that I could do more complex stuff with more channels.

This static you are getting. Is it constant? Do you get it in normal 4-channel mode? If so, the problem might be ageing capacitors.

If you only get it in mix mode, have you checked whether or not the "echo" effect is turned on? It has a tendency not to properly fade to zero, causing a  small amount of repeat noise after a sample has played. Hitting the space bar cancels all sample playback.

As a test, load a regular 4 channel mod. Confirm that it plays correctly in regular 4-channel mode. If so, change to Mix mode in the song options and set your mixing settings to 4 channel mono, no smoothing, 22kHz. Go to the Edit->Transpose menu and bump everything up by 2 octaves.

Now play the track again. If all is well, you should hear it playing back in mono and it shouldn't completely soak up your CPU. Now switch to stereo and try again.

Adding fast ram alone makes a big difference to octamed's mix mode. This is because the software mixing routines aren't held up reading the sample data from chip ram, they can read it straight from fast ram. This, in conjunction with the general speed up of having fast ram makes a huge difference.

Once you get your 030 with fast ram, you should be able to use up to 16 channels at 27kHz. With an 040, I was able to use over 24. I don't think I actually wrote any pure (non midi) module that needed more than that

[Karlos]

Karlos is my hero, great posts...

Steady on, feller :lol:

I didn't know you were down with the music stuff so hardcore! I got some midi questions/problems I might bother you about.

Somewhere in the depths of this forum, I believe it's a matter of public record that one of my main interests in the amiga was making music with it. The first apps I got for my new A600 were Protracker and OctaMED. The first peripherals I bought for it were a parallel port sampler (StereoMaster, if anybody remembers that) and a MIDI interface to control my old Yamaha PSS680, my first multitimbral MIDI instrument.

I got my first A1200 10 months after that initial A600 to be able to do more complex musical tasks (gaming was nice but a bit secondary at the time), knowing that it was more expandable.

I got my first hard disk (aside from the general convenience) so I wouldn't have to rely on an increasingly large library of floppy disks for samples, my first ram expansion to make the thing more productive. I got my first accelerator (an 040 no less) so that I could make better use of OctaMED SS at the time. From the moment I got the demo version, I knew it was my "perfect app".

I've used my Amigas for a lot more than that, of course, as you do any computer system. Coding became a major interest too. However, having had most of the other reasons for having a computer usurped by more modern systems, music remains the one computer hobby that has not migrated to the PC. The only thing I use my PC for in that respect is audio capture.

[Karlos]

Awesome, you know, think I understand the synth editor now and made my 1st synth script this past weekend, I'll try and share that, it is a pretty cool and underated feature of OSS.

If you want to play with realtime synthesis, you must give MusicLineEditor a go.