Combining excerpts of several sessions

Hi everyone. I am a brand new owner of a Kasina, and I wrote to Andrew in support with some unusual questions about combining short excerpts of several Kasina sessions, similar to what Saswing did with his spectrastrobe sampler here: http://www.mindplacesupport.com/foru...ectra-sessions. I will paste in my questions and Andrews answers below in case any one has some further ideas on how to go about this:

Here is what I wrote to Andrew:

I really like the track "Audio Illusions" in the Audiostrobe folder because, as I mentioned, it has a lot of musical interest, mostly because the music changes a lot as the track plays. I was thinking of making my own "mix tapes" made up of short segments of the various tracks on the Kasina, as well as tracks I have downloaded from the forum and soundcloud. So for example, I could start with 20-30 seconds of a quiet peaceful piece from the meditate category, and then segue into 20-30 seconds of a more upbeat piece from the mind art category, followed by 30 seconds of one of the pieces I downloaded from the forums, and so on. I would use an audio editor like Audacity to combine the excerpts from several tracks into a single mp3.

So my questions are:
1- If I take shorter excerpts from the Spectrastrobe tracks, will the spectrastrobe encoding still work to control the lights? Or is the spectrastrobe encoding located all at the beginning of the complete (long) tracks?
2-If the spectrastrobe encoding is distributed throughout each track, then do you think I can overlap two Spectrastrobe tracks....i.e. fade out the older track and fade in the newer track kind of like how a DJ does it at a dance party? Or will that confuse the Kasina to have two sets of spectrastrobe instructions playing at the same time? If I need to, I could just fade the first track out completely before fading in the second track. But it could be nice to have the tracks overlap for more continuity.
3- And what about combining Audiostrobe and Spectrastrobe excerpts in the same "mix tape"? Will the Kasina be able to sense the change of encoding and automatically switch from say audiostrobe to spectrastrobe and back?

I will completely understand if you do not know the answers, as these are not the usual kind of questions. I just find that after a minute or two of peaceful forest sounds, I am ready for something different (even just a slight change to ocean sounds or binaural beats can help keep it from being monotonous). Again, that is why I like the Audio Illusions track because the musical content shifts around a lot. I am hoping I can create my own combination tracks using the sessions that have already been created for the Kasina. Thanks for taking the time with questions from a new user of the Kasina!

And here is Andrew's reply:

I'm glad to hear that you were able to download some of the free tracks and are looking to expand your collection.

To answer your questions:
1. SpectraStrobe and AudioStrobe are encodings that happen in real-time within the tracks. Editing the sessions should not affect the existing encoding signals, however, I would highly recommend that you use an editor which can edit the .mp3 files directly without decompressing and recompressing. The de-compression/re-compression procedure may introduce anomolies in the encoding, or reduce the quality of the encoding.
If you must decompress, then do it to a .wav file and do not re-encode the file to .mp3 after editing. The file will be bigger, but you won't lose any quality or encoding that way.

2. You should give it a try and see how it works. Best way to find out is by experimenting.

I'm sure others would be very interested in your findings and perhaps others have done the same thing. Why not register at the forums and discuss what you're doing? Others may be able to suggest ideas and/or learn from what you are doing. I think it's a great topic for the forum.

Enjoy!

If anyone has any more ideas or suggestions, please chime in. Thanks!

I also found this blog post that describes how to make a playlist by creating a new folder on the kasina:
http://www.mindplacesupport.com/foru...ists-on-Kasina

It might be a lot quicker and less "fussy" to just create a bunch of short excerpts of mp3s that fade in and out, and then put them all in a folder. That way they could all play in sequence, but I would not have to get the files all combined into one mp3.

I may try this method first as it seems simpler to me. I found this program for cutting the files without re-encoding: http://mpesch3.de1.cc/mp3dc.html

Edit added: I just tried out the mp3DirectCut program and it worked fine to make some short excerpts from my various sessions. It even has a feature that allows you to put a fade in or out at the beginning and end of tracks. (One tip: use the right hand mouse button to make selections in this program)

I put all of the tracks in a folder and will try them out later today as a "playlist". I made most selections at least 3 minutes long so that it would not jump around too much.

I could put a zip file on here of the edited files, but I am not sure if that is really kosher since some of them come from albums I purchased. Would it be OK to attach a zip file made up only of shorter versions of sessions that are already on the Kasina? If anyone knows if that is OK, let me know and I will gladly share my edited (i.e. shorter) versions on here.

I have one discovery to report. It seems you need to include the beginning of a spectrastrobe session. If I cut the beginning off of the session, then the Kasina displays an error message and the session does not play.

Otherwise, the play list idea seems to work.

Hello Nirmala,
Your experiments are cool. Some of the shorter sessions on the Kasina are actually designed to be used in this way. That is combining them with together and with other 'bumper'sessions. This is an excellent way to get a great deal of variety from the sessions. We will expand on this concept more.
Still what you are doing was not anticipated and is creative. I am interested to hear what combinations you come up with. One thing of note also is that you can mix different session types in the same folder, AudioStrobe, SpectraStrobe,KBS and regular music files. Each one has a different set of 'features' so to speak. So you could do a standard music intro that you like using the ColorOrgan, which the switches to some complex visuals with AudioStrobe using the ColorSets, then to a KBS file that delivers a basic entrainment section, then end on a high note with some trippy SpectraStrobe. One example of what is possible.

Now on your report. There is no special encoding at the beginning of a SpectraStrobe track. You can pick it up at any point. It might have slight glitches at the transition point where the SS reference tracks fades in or out, but otherwise it should work.
I suspect that the software you are using for the cutting is putting in non-standard headers in the mp3. We have run into this before and Andy can elaborate.
Or it could be that the settings of the mp3 output are being altered. The recommend encode settings are 256kbps or above, constant bit rate CBR, with normal stereo separation. You could check that out to see if something is changing.
As I said, there is no technical reason you couldn't pick up the SpectraStrobe or AudioStrobe from any point in the file.

I am interested in your results and thanks for sharing your experiments.

Scott

Thanks Scott for your ideas and perspectives. It is nice to be able to combine different types of sessions. And it is very easy with the mp3directcut program to make shorter sessions out of the existing sessions on the Kasina.

Another possibility as to why the Kasina shows an error message and won't play the tracks I created where I cut off the beginning of a session is that I used the mp3directcut program to also make a short fade-in at the start of the edited excerpt (the shorter sessions that include the beginning of the original usually already have a fade-in that is part of the original). Perhaps the code used to create the fade in messes up the spectrastrobe encoding in the fade-in section, and that is what confuses the Kasina. I will try again making a short session from the middle of an existing session, but without applying the fade-in feature on the mp3directcut software.

I did use the fade feature to create gentle fade-outs at the end of the short sessions, and that seems to work fine. The mp3directcut software is not supposed to do any decoding or encoding of the file and is supposed to maintain the original bit rate. I just checked and the edited spectrastrobe files are still 320kbps.

Hello,
A mp3 has a header at the beginning that stores all the metadata (artist, title,etc) In this header there are values pointing to the 'data' that is where the music starts and also the track length, also where the 'data' ends. Other stuff is in there too. I think your mp3cutter program is maybe not updating these headers correctly. Or some other form of non-standard header alteration. It is not necessarily changing the encoding but it must change these header values.

The Kasina itself should only display errors and refuse to play the file if the onboard mp3 codec doesn't recognize the file as a standard mp3. SpectraStrobe processing is later in the chain. Errors with it would only really show up as SpectraStr/AS being grayed out or sensing AS when you know the file is SS. Those appear on the Kasina session display when the onboard audio file or AUX input source is playing. With problems with SS, you would not get the error message you have described. This indicates a problem with how the audio file is being recognized.

Scott

Just a suggestion but since you are getting into cutting and mixing, you might want to spend some time learning the basics of Audacity, if you haven't before. It is free and powerful. It would do this process more cleanly. You will have to also install the Audacity Lame mp3 plugin. When you open a factory SS file in Audacity it makes a temporary wav copy of the file and then you can work with it and once done export it back to mp3. As long as you keep the bitrates and don't keep reworking the same file over and over again (meaning recompressing it) then you should be ok. Lame is a great mp3 encoder and used in all the factory stuff.

Possibly then it is the fade-in feature that is changing the headers. I am pretty familiar with Audacity, but I did not use it because of what Andrew said about not using an editor that decompresses and recompresses the file. It took a while to find a simple editor that does not do that, but then I found the mp3directcut software.

The edited files with the fade-in at the beginning play fine on my computer, so I am not sure why the Kasina does not recognize them as mp3s, again unless the fade-in changes the headers. It is not a big issue, as the files work fine if there is no added fade-in.

Computer software is much more forgiving than hardware. Software has all kinds of conditionals that take actions when unexpected things occurs. Your PC has a lot more memory space to contain this complex code. In addition the makers of the software tested it against the common mp3 programs and not our hardware.

Not sure about the fade in either. It couldn't actually do a fade in without altering the audio content, at least I don't think so. But maybe there is a way to place that instruction in the header. I don't know about it though.

As you say in the end you've got an acceptable solution. We are always curious about what might cause errors. And also it is cool when users discover new tricks.

I agree that it's important to not decompress and recompress the mp3s. They've already lost some of their data in the initial conversion from .wav to .mp3, and to do the conversion again will lower the quality further or possibly cause issues with the spectrastrobe/audiostrobe encoding. Best to do the cutting with a program that doesn't decompress/recompress as you are doing.

Can you make a short clip that won't play on the Kasina and attach it here so I can see what is wrong with the file?

Still as mentioned the only way that this or any software can trim without recompressing is by manipulating the header values and truncating the file. We have not tested against this internally. But I suppose we should.

According to the wikipedia entry on this software it is manipulating the headers and frames to make the changes. http://en.wikipedia.org/wiki/Mp3DirectCut

Also this from that Wikipedia entry:
Gains and fades are limited to 1.5 dB steps, as they are achieved by modifying the value of the MP3 frame's 8-bit global gain field.

I suspect this is putting in spurious values as far as Kasina is concerned. Or there is a checksum mismatch.

Here is one edited session that would not play...

Andy is of course correct when he says that repeated conversions between MP3 and lossless formats like WAV will tend to degrade the signal quality over time. But a single cycle might not matter much: I recently encoded some MP3s with SS for a customer, and used Audacity to convert to WAV before importing to Mind Workstation. When finished I could not hear any noticeable degradation of the audio signal in the final MP3. I did notice though that if I allowed the SS signal to continue past the end of the audio track, then faded that signal out, the lights did not accordingly fade, but instead the Kasina lost the ability to track those signals and the lights actually grew brighter and stopped flashing. Fading the brightness in MWS works like a charm, as expected. Aside from the header and other issues arising from directly editing the MP3, as mentioned by Andy and Scott, fading in and out a section of the SS encoded files will likely introduce this sort of visual glitch at the boundaries between excerpts...

-Robert

Quote:

---

Originally Posted by Nirmala

Here is one edited session that would not play...

---

This track plays on my Kasina. Do you have the latest firmware on your Kasina?
1.10 is the version available on the support site.

Quote:

---

Originally Posted by Andy

This track plays on my Kasina. Do you have the latest firmware on your Kasina?
1.10 is the version available on the support site.

---

I am busy the next several days, but I will check into this next week. And I will double check the file on my Kasina....maybe the file just got corrupted when I was loading it onto the SD card, but there were 2 other files that would not play also.

Ok, my first guess is that you have the original firmware on there and updating to 1.10 will fix the problem.

Let me know what you find.