Own Voice Recording

nix · 19 Jun 2009

oddson wrote: I'm lost... what does all this float array and green trigger stuff have to do with recording/saving audio? 😕

Float arrays equate to mems if you want them to. My project that you've seen in progress uses float arrays to make cuts to the wave. I thought that one way to record would be to feed a tick44100 to a float array? Tzarls solution is great though. The tick1000 isn't really in the right topic, it just came up while I was experimenting. @AM 😍

Tzarls · 21 Jun 2009

I was thinking about this tick1000 thing, and came to a extremely simple idea. I have built a tick5X. If you attach a tick100 to it, it will become a tick500. No red lines, no CPU stall (but is CPU intensive, though).

nix · 21 Jun 2009

Nice work! Much thanks, btw it uses no cpu on my quad3 to make tick 500. I'll ramp it up and see what happens. Great for an array renderer! It ticks at increments of one because of the link order? edit ps- tick 2000 uses 25% edit2- Woah! I can render waves at tick2000 rate as long as the climbing no. of entries isn't displayed without impacting cpu I think! I'll try 44100hz and the renderer can keep up with samplerate near enough. Still the timing issues though. I don't know what this is for but it's pretty cool though! :})

Disco_Steve · 21 Jun 2009

sorry but for me the tick *5 just makes the normal tick 4 times slower 🙁

aliasant · 21 Jun 2009

Disco_Steve wrote: sorry but for me the tick *5 just makes the normal tick 4 times slower 🙁

Huh. WEird. Worked excellent here. I tried several config changes and all worked. So surprised that it was so simple to change the trigger.

Tried your osm there Steve. Sort of works. The lower one gives me 5x ticks but isnt it kind of interesting that, for me at least, the 100 during a time of 1000ms only outputs 64 ticks? Over and over again. Occasionally 65 ticks.

Disco_Steve · 21 Jun 2009

I get 24 every time the *5 slows down the normal tick by 4 for me. which makes sense as the *5 is made by adding an extra 4 triggers stacks 'fuelled' from the first. so for the rate to go down by 4 makes lots of sense. but for you to get 64? wha? that doesn't make sense to me? what computer do you have? on my 'new' machine core2 quad the green seems to run at a very consistent 100 ticks per second always. but on the p4 down stares the green works much faster. very bizar? what numbers do other people get?

aliasant · 21 Jun 2009

Ok. Image proof. Heres my result after 1 click. And I remembered to reset the counters first 😉

I have a Mac Pro with 2 dual xeons 2.66. Total of 4 cores. Running Win7.

Disco_Steve · 21 Jun 2009

can I have your machine please 🙂 I did believe you just weired out more really by the consistency in my machine. I don't think I bought the right motherboard.

aliasant · 21 Jun 2009

Disco_Steve wrote: can I have your machine please 🙂 I did believe you just weired out more really by the consistency in my machine. I don't think I bought the right motherboard.

Oops. Sure. If you finish all my plugins, optimize them, and fix a website that sells them =) I would be happy to give it to you.

But seriously. Cant you just get a new motherboard? Buy some quality stuff from Asus. Seems to be a pretty safe bet.

Would be interesting to hear other peoples result. I didnt know that the tick was so sensitive to what theyr running on. I thought a Tick100 would tick a hundred times give or take a few just because its windows...... lol

nix · 22 Jun 2009

Here is the array renderer(recorder,writer) as well as can work with the green I think. It has a tick 44100, if anyone needs that(sorry windows timer). It's pretty hefty on cpu, make sure and turn off the bool to the save wave when recording I think. This is what AM had in mind sorta. I'll have another crack at the timing with tzarls samplerate stream counter posted in this thread sampled to green at high hz.

Bids content PS- yeah the timers go slowly sometimes when modding on my system. You have to be careful when displaying things maybe.

aliasant · 22 Jun 2009

I did a test with this 5xtick and I dont really think it works? At least not for me.

I used a counter and fed the ticks into that one. I thought that if it sends 5 ticks it should count up to 5, gradually. 1. 2. 3. 4. 5 But it simply jumped directly to 5

Then I tried using it in a loop. I had 2 identical loops with GetAt and Array but on one of them I added the 5xTick just after the loop but before the GetAt.

This took much longer time to calculate. i would guess 5 times longer...eh. The result was the same except that it took 5 times longer and used 5 times the amount of ticks.

I had a look at nix thing but I dont quite understand it. A bit to tired too.

I dunno what to think.

Tzarls · 22 Jun 2009

Here´s an improved version of the 5x ticker, it uses no int sample and holds (no memory usage), only triggers. I made a 125x just as an example. I think there are 2 problems here. The tick 100 (by itself, with the 5x ticker disabled) puts out about 65 ticks per second, not 100. Weird. The second problem is that you can´t have on display anything related to the ticks (like a counter displaying in realtime how many ticks are being sent) because displays slow down everything.

Now, Aliasant, I don´t understand what you mean with your "counter" example. If the counter jumps to 5, it means it received 5 ticks, right?

aliasant · 22 Jun 2009

Tzarls wrote: Here´s an improved version of the 5x ticker, it uses no int sample and holds (no memory usage), only triggers. I made a 125x just as an example. I think there are 2 problems here. The tick 100 (by itself, with the 5x ticker disabled) puts out about 65 ticks per second, not 100. Weird. The second problem is that you can´t have on display anything related to the ticks (like a counter displaying in realtime how many ticks are being sent) because displays slow down everything.

Now, Aliasant, I don´t understand what you mean with your "counter" example. If the counter jumps to 5, it means it received 5 ticks, right?

Well I also had the counter hooked up to a GetAt and an Array and I didnt get the results I was hoping for. Do some real world tests yourself. If you can make it actually work thats very cool but I doubt it really works

Tzarls · 22 Jun 2009

Problem here is that I don´t know what results you are expecting. And I didn´t build this for some specific application. In fact, I only tried to improve the "tick1000" posted in this thread. For me, this only proves that we can send triggers very fast!(faster than the tick100, at least) If useful, I really don´t know.