| 00:21.22 | *** join/#utah linuxalien (i=linuxali@gateway/tor/x-fa37ffab437ca168) |
| 00:25.44 | vontrapp | Sargun: blender? |
| 00:26.17 | Sargun | mhmm |
| 00:27.28 | vontrapp | atoponce: etherape |
| 00:31.45 | *** join/#utah elg (n=fugalh@falcon.fugal.net) |
| 00:31.53 | elg | dilvie: ping |
| 00:33.51 | dilvie | pong |
| 00:34.36 | dilvie | elg: get my email? I've got tortoisedarcs, but I'm not sure how to connect.. do I need a username? |
| 00:35.06 | dilvie | looks like it should be a breeze, if I get the path and login correct. =) |
| 00:44.54 | elg | yeah, I just replied, and sent you one with detuning, as it is |
| 00:45.03 | elg | just give darcs that http url |
| 00:45.05 | dilvie | yay |
| 00:45.15 | dilvie | no login? |
| 00:45.21 | elg | caveat given in the email |
| 00:45.33 | elg | no, not needed. of course you won't be able to commit anything without a login |
| 00:45.48 | elg | but you can prepare a patch real easy-like and send it to me |
| 00:45.55 | elg | and I can apply it real easy-like |
| 00:46.08 | atoponce | vontrapp: not what i'm looking for i think. how do you show running stats and port numbers? |
| 00:46.24 | elg | with the cli version, it's "darcs send" |
| 00:46.54 | elg | note that darcs is a distributed rcs, so you do commit, you just commit locally |
| 00:47.13 | elg | then you send those patchsets you wish to send along to me |
| 00:47.23 | dilvie | cool |
| 00:48.15 | elg | do you see the issue with the detuning? |
| 00:48.25 | vontrapp | atoponce: in etherape? click the prot. graph chart button |
| 00:49.48 | dilvie | elg: yeah, I see it. How do autotune plugins work? |
| 00:49.58 | elg | got me |
| 00:50.27 | elg | there's probably some fancy tricks that can be done. basically you'd have to fudge something |
| 00:50.34 | elg | you can't just slow down the sample rate or you get behind |
| 00:50.50 | elg | so you've got to resample and stretch somehow, and do it in realtime |
| 00:51.08 | elg | if you find a good algorithm, I'll implement it. :) |
| 00:53.07 | dilvie | realtime sample rate conversion? |
| 00:53.52 | elg | well resampling in realtime is easy enough |
| 00:54.06 | elg | it's keeping it on the same timescale that's hard |
| 00:54.33 | elg | it's not time-invariant, to use a term I just learned in my dsp class :) |
| 00:56.03 | elg | what you want is pitch scaling without changing to duration |
| 00:56.23 | dilvie | http://www.tobybear.de/dsp.htm |
| 00:56.25 | atoponce | vontrapp: ok, i see running stats with each protocol, but how can i see the amount of traffic on each port? |
| 00:56.36 | vontrapp | shrug |
| 00:56.52 | dilvie | I'm looking for a pitch scaling algo |
| 00:57.46 | dilvie | http://www.dspdimension.com/data/index.html |
| 00:58.51 | elg | I've seen that one. I'll read it again |
| 01:00.11 | dilvie | here I thought the detune would be easy. |
| 01:00.15 | dilvie | =P |
| 01:00.43 | dilvie | you saw the tutorials? |
| 01:00.45 | dilvie | http://www.dspdimension.com/data/index.html |
| 01:01.30 | dilvie | http://www.dspdimension.com/data/index.html |
| 01:01.36 | dilvie | ur, woops |
| 01:01.42 | dilvie | sorry, he's using fram.s |
| 01:01.44 | dilvie | frames |
| 01:01.59 | dilvie | check tutorials -> pitch shifting code |
| 01:02.15 | elg | this is the one I tried to implement last semester |
| 01:02.16 | elg | http://www.csounds.com/ezine/summer2000/processing/ |
| 01:02.24 | elg | sounded terrible |
| 01:04.32 | elg | http://en.wikipedia.org/wiki/Audio_timescale-pitch_modification |
| 01:06.24 | elg | this looks good http://www.dspdimension.com/data/html/timepitch.html |
| 01:07.23 | vontrapp | atoponce: did you try tcp mode? |
| 01:07.27 | *** join/#utah TacoJockey (i=linuxali@gateway/tor/x-50b878c9fb672e6d) |
| 01:08.08 | elg | i think part of why that one sounded bad is that it's frequency shifting, and not musical |
| 01:11.28 | dilvie | hmm |
| 01:11.42 | atoponce | vontrapp: yeah. it's not what i'm looking for. i think darkstat and iftraf will do fine |
| 01:11.52 | dilvie | elg: we're working at such a small shift, I don't think we should worry about formants |
| 01:11.58 | elg | listen to this http://www.dspdimension.com/data/audio/FSHIFT.WAV |
| 01:12.02 | elg | yeah, formants aren't an issue |
| 01:12.29 | dilvie | formant shifts only seem to pose a problem when you shift by more than 3 whole steps (400 cents) |
| 01:12.32 | elg | btw, listen to the one I sent you - it might just be good enough (sometimes) |
| 01:13.05 | elg | and this http://www.dspdimension.com/data/audio/PSHIFT.WAV |
| 01:13.30 | elg | and the original http://www.dspdimension.com/data/audio/ORIG1.WAV |
| 01:14.06 | dilvie | that fshift seems to be just a formant shift effect |
| 01:14.12 | dilvie | that's spooky cool. |
| 01:14.18 | dilvie | I'd love a formant shift plugin! =) |
| 01:14.24 | elg | hehe |
| 01:14.44 | dilvie | that's a different plugin, though |
| 01:14.51 | dilvie | I don't think we need to worry about it in a haas plugin. |
| 01:14.53 | dilvie | =) |
| 01:15.13 | elg | the middle one (PSHIFT) sounds ok, esp I assume at the 15 cent range |
| 01:15.43 | elg | that phase vocoder algo is complicated though |
| 01:15.50 | elg | might be better off as its own plugin |
| 01:19.03 | dilvie | I'm going to test the current pitch shifting |
| 01:22.00 | dilvie | hmm |
| 01:22.05 | dilvie | seems 15 cents isn't enough |
| 01:24.28 | dilvie | I thought I tested the range when I was drawing up the specs, but I guess I didn't ); |
| 01:24.47 | elg | easy to change |
| 01:24.54 | dilvie | maybe we should bump it up to 100 cents |
| 01:26.21 | dilvie | that will give us more than enough range to create valid haas effects |
| 01:26.37 | dilvie | I've actually used 100 cent offsets for drums, to good effect |
| 01:26.58 | dilvie | more than 100 gets weird, though, even on drums. |
| 01:27.40 | dilvie | and can be accomplished by using different notes and MIDI pitch shifts |
| 01:29.17 | *** join/#utah linuxalien (i=linuxali@gateway/tor/x-81d55fd3cd5062f9) |
| 01:32.45 | elg | dilvie: http://www56.homepage.villanova.edu/scott.sawyer/fpga/ |
| 01:33.31 | elg | I think that's got everything I need. I'll be getting up to speed in matlab the next week or two, I'll run their matlab code on some music (feel free to send me some sample of what you want to shift) at various numbers of cents |
| 01:33.58 | elg | and then we can decide which one to implement. that hilbert transform one doesn't look too bad - I'm sure we'll be covering hilbert transform in class too |
| 01:34.22 | elg | it's very likely that getting the perfect detuner will take another semester, though. |
| 01:36.16 | dilvie | =) |
| 01:36.23 | dilvie | Any chance you can use this for credit? |
| 01:38.44 | elg | i did, last semester |
| 01:39.15 | dilvie | cool. |
| 01:40.08 | dilvie | can you send a version with a 100 cent range? |
| 01:41.36 | elg | yeah |
| 01:44.36 | elg | snet |
| 01:44.47 | dilvie | awesome. |
| 01:45.41 | dilvie | would you object to moving this project to sourceforge/svn, once we've got it to a releaseable state? |
| 01:46.10 | dilvie | there are a few other developers who are interested in it. |
| 01:46.18 | elg | not if you want to take it over. :) |
| 01:46.33 | elg | we can work something out, for sure. I'm not much a fan of sf, though |
| 01:46.44 | elg | and would prefer darcs |
| 01:46.52 | dilvie | cool.. I want to bring it to a usable state, first, and then I'll put it up on sf and add the hpf =) |
| 01:46.55 | elg | but if someone else wants to take the reins they're welcome to do it however |
| 01:47.25 | elg | I'm trying to remember if I used any gpl code |
| 01:47.41 | dilvie | well, the cool thing about sf, is that we can have a couple different project admins, and if somebody wants to move on to other things, the community can pick it up and keep supporting it. |
| 01:47.53 | elg | i think it's all mine. so we could license it with a bsd license or something compatible with vst's license so we can distribute binaries |
| 01:48.09 | dilvie | cool. =) |
| 01:48.13 | elg | maybe this is a good chance to try out google's code hosting service |
| 01:48.25 | dilvie | oh, that might be cool. =) |
| 01:51.47 | dilvie | detune isn't changing the pitch |
| 01:52.33 | dilvie | maybe it is |
| 01:54.24 | dilvie | elg: it's not changing the pitch, it seems to be changing formants only? |
| 01:57.12 | dilvie | It's really cool hearing the sidebands shifting as I move the slider, but that's not what this setting is supposed to do. ;) |
| 01:59.54 | elg | neither |
| 02:00.07 | elg | the buffer is too small |
| 02:00.16 | elg | i'll send you yet another :) |
| 02:01.22 | elg | do you have a build env yet? |
| 02:02.15 | dilvie | yeah |
| 02:03.36 | dilvie | elg: it was changing formants.. I'm not sure why, but the high side-band frequencies were shifting up and down when I moved the slider back and forth. |
| 02:04.00 | elg | probably because the high frequencies were the ones that could be effected in that small buffer |
| 02:04.26 | dilvie | cool |
| 02:05.46 | dilvie | it needs to be able to pitch-shift a sine wave down to 20hz |
| 02:06.51 | dilvie | when I get soundforge installed on this thing, I'll test all the specs and make sure that all the settings are accurate |
| 02:07.20 | dilvie | I'm using a brand-new workstation, and I haven't installed all my software, yet. |
| 02:09.40 | elg | a 440 sine wave? |
| 02:09.46 | elg | I think you'll like this one |
| 02:10.27 | dilvie | uh, no |
| 02:10.46 | dilvie | a sine wave at 20hz + 100 cents should be able to shift down to 20hz |
| 02:10.48 | dilvie | =) |
| 02:11.34 | dilvie | 440hz down to 20hz would be a lot more than a 100 cent shift. ;) |
| 02:12.12 | dilvie | 100 cents = 1/12 octave. a doubling in frequency = 1 octave |
| 02:13.06 | elg | 440 to 415 is 100 cents |
| 02:13.14 | dilvie | so 220hz - 440hz = 1200 cents |
| 02:13.31 | elg | oh, you said down to |
| 02:13.37 | elg | I thought you meant shift by 20hz |
| 02:13.45 | dilvie | no |
| 02:14.26 | elg | let's see. how many samples does it take to shift 20 down to 18.87 |
| 02:14.28 | dilvie | I was just making a point that the buffer needs to support the whole audio frequency spectrum |
| 02:14.53 | dilvie | audible spectrum (in general) is 20hz - 20khz |
| 02:15.25 | elg | my brain is shot or I'd do the math |
| 02:15.48 | dilvie | well, 10hz - 20hz = 1200 cents |
| 02:15.50 | elg | you can fiddle with the #define. set it to 48000 and see what happens if it gets too long |
| 02:16.12 | elg | then try to find the smallest one that covers 20hz |
| 02:16.34 | elg | the macro is DETUNE_RING_m |
| 02:16.48 | elg | in common/detune.h |
| 02:18.43 | dilvie | 20hz - 120 cents = 19hz, right? |
| 02:19.12 | dilvie | probably not right |
| 02:19.12 | elg | 20*(2^(120/1200)) |
| 02:19.21 | elg | oops, 20 / (2^(120/1200)) |
| 02:19.22 | dilvie | I suck at this stuff. =) |
| 02:19.26 | elg | * is shifting up |
| 02:19.46 | elg | 18.66 |
| 02:20.07 | elg | s/shifting/scaling/ |
| 02:21.07 | dilvie | I'll test the current setting and let you know if there's a problem. =) |
| 02:21.36 | elg | ok, if I'm doing the math right, you need a period of 3200 samples to cover 20hz |
| 02:22.01 | elg | which corresponds to .066 seconds (at 48khz) |
| 02:22.12 | elg | actually it's the time that matters, and the samples relates to that |
| 02:22.26 | elg | but that will introduce a very nasty delay |
| 02:22.44 | dilvie | uh-oh |
| 02:22.49 | dilvie | hrm. |
| 02:22.51 | elg | so I recommend we let the lower frequencies fly and fudge it. :) |
| 02:23.11 | dilvie | is 50hz doable? |
| 02:23.28 | elg | .02 seconds |
| 02:23.30 | elg | 1/50 |
| 02:24.14 | elg | i might be wrong that you need a whole period for the frequency to be scaled |
| 02:24.19 | dilvie | wow, this realtime audio dsp stuff is complicated. =) |
| 02:24.35 | elg | we could make the length of that buffer a parameter |
| 02:24.41 | elg | at least while we find a good value |
| 02:24.56 | dilvie | that's probably a good idea, for testing purposes. =) |
| 02:24.58 | elg | but for now just change that value and recompile |
| 02:27.53 | dilvie | the current value is still not even close to usable |
| 02:29.09 | *** join/#utah drewbono (n=andrew@pool-71-126-132-147.washdc.fios.verizon.net) |
| 02:30.02 | dilvie | it needs to be able to go down to at least 100hz |
| 02:30.38 | dilvie | Ideally, 20hz. 50hz would be nice. =) |
| 02:30.58 | dilvie | maybe we need to find a different algo.. |
| 02:41.39 | Sargun | what are you guys doing? |
| 02:46.58 | dilvie | Sargun: working on an audio effect plugin that does panning that your surround sound decoder can understand. =) |
| 02:47.06 | Sargun | sweet |
| 02:47.08 | Sargun | Sweet |
| 02:47.13 | Sargun | Mucho Sweeto |
| 02:47.26 | Sargun | plug-in to what? |
| 02:57.04 | *** join/#utah drewbono (n=andrew@pool-71-126-132-147.washdc.fios.verizon.net) |
| 03:36.44 | elg | a vst plugin |
| 03:38.44 | elg | and ladspa |
| 04:01.08 | *** part/#utah linuxalien (i=linuxali@gateway/tor/x-81d55fd3cd5062f9) |
| 05:05.12 | *** join/#utah herlo (n=herlo@71-219-25-112.slkc.qwest.net) |
| 05:21.14 | *** join/#utah herlo (n=herlo@71-219-25-112.slkc.qwest.net) |
| 05:23.01 | *** join/#utah linuxalien (i=linuxali@gateway/tor/x-e1844a1a74cccbe7) |
| 05:24.11 | *** join/#utah vontrapp (n=von@71.35.254.57) |
| 05:26.39 | *** part/#utah linuxalien (i=linuxali@gateway/tor/x-e1844a1a74cccbe7) |
| 06:18.05 | *** join/#utah hatchmt (n=mhatch@71-219-21-234.slkc.qwest.net) |
| 07:42.55 | Sargun | Anyone here have a Laser? |
| 07:52.43 | Tene | Yay, my laptop is broken! |
| 09:21.20 | Sargun | :-/ |
| 09:21.22 | Sargun | what happened? |
| 12:14.32 | *** join/#utah linuxalien (i=linuxali@gateway/tor/x-3aef5d62e5fb9bd7) |
| 14:38.50 | *** join/#utah elg (n=fugalh@falcon.fugal.net) |
| 15:46.40 | *** join/#utah neybar (n=jalance@thelances.net) |
| 15:56.59 | linuxalien | ~lart ibot |
| 15:57.16 | linuxalien | You tart |
| 16:06.18 | ScytheBlade1 | haha |
| 16:49.50 | *** join/#utah jnbek (n=jnbek@c-24-10-178-17.hsd1.ut.comcast.net) |
| 16:50.16 | *** join/#utah redbeard2 (n=redbeard@c-67-182-237-149.hsd1.ut.comcast.net) |
| 17:15.34 | *** join/#utah sontek_ (n=sontek@c-67-182-254-162.hsd1.ut.comcast.net) |
| 18:35.31 | *** join/#utah vontrapp_ (n=von@71.35.254.57) |
| 18:55.53 | *** join/#utah hatchmt (n=mhatch@71-219-21-234.slkc.qwest.net) |
| 18:58.41 | *** part/#utah linuxalien (i=linuxali@gateway/tor/x-3aef5d62e5fb9bd7) |
| 19:18.30 | *** join/#utah dilvie (n=chatzill@c-67-166-126-133.hsd1.ut.comcast.net) |
| 19:25.14 | *** join/#utah vontrapp (n=von@71.35.254.57) |
| 19:40.27 | *** join/#utah linuxalien (i=linuxali@gateway/tor/x-c94a28509a4c7561) |
| 19:57.31 | *** join/#utah herlo (n=herlo@71-219-25-112.slkc.qwest.net) |
| 20:00.09 | dilvie | you guys are a lively bunch today! |
| 20:02.12 | linuxalien | no we're not. Liar. |
| 20:04.37 | herlo | linuxalien: yes we are, just not in here, bugger off! |
| 20:07.00 | linuxalien | hehehe :) |
| 20:07.03 | linuxalien | gotta run |
| 20:07.07 | linuxalien | gonna pick up my brother |
| 20:20.38 | *** join/#utah dilvie (n=chatzill@c-67-166-126-133.hsd1.ut.comcast.net) |
| 21:00.56 | *** join/#utah markl_ (n=mark@dsl093-225-013.slc1.dsl.speakeasy.net) |
| 21:08.55 | *** join/#utah elg (n=fugalh@falcon.fugal.net) |
| 21:20.47 | *** join/#utah Specialk (n=mse@128.187.161.235) |
| 21:24.31 | Specialk | ~lart maquis |
| 21:36.12 | *** join/#utah dilvie (n=chatzill@c-67-166-126-133.hsd1.ut.comcast.net) |
| 22:16.17 | *** join/#utah herlo (n=herlo@71-219-25-112.slkc.qwest.net) |
| 22:25.12 | *** join/#utah levi_home (n=levi@levi.dsl.xmission.com) |
| 22:47.40 | *** join/#utah Lone_Wanderer (n=chatzill@c-67-177-37-53.hsd1.co.comcast.net) |
| 23:01.29 | Lone_Wanderer | Is there somewhere I can read about different distros and the pros/cons/personalities of each? |
| 23:04.11 | findlay | distrowatch |
| 23:04.22 | Lone_Wanderer | thanks |
| 23:07.07 | maquis | -lart specialk |
| 23:07.20 | Lone_Wanderer | ~lart specialk |
| 23:07.28 | Lone_Wanderer | You're welcome, darling. |
| 23:07.37 | maquis | Lone_Wanderer: thx |
| 23:07.59 | Lone_Wanderer | You mean the cereal? |
| 23:09.07 | maquis | i meant specialk who came in and larted me, then left |
| 23:10.51 | Lone_Wanderer | oh :( |
| 23:10.52 | Lone_Wanderer | meanie head |
| 23:12.29 | Tene | ~levi maquis |
| 23:12.30 | ibot | ACTION gives maquis a hovercraft and then fills it with eels! |
| 23:13.41 | Lone_Wanderer | is ~levi related to Levi Pearson? |
| 23:14.25 | Tene | ~whois levi |
| 23:14.38 | Tene | yeah |
| 23:16.13 | Lone_Wanderer | That confuses me, but okay. |
| 23:17.11 | Lone_Wanderer | "FORTRAN was the language of choice for the same reason that three-legged races are popular." |
| 23:52.42 | *** join/#utah Sargun (n=xbmodder@69-12-240-206.dsl.dynamic.sonic.net) |