| *** rpierce99 has quit IRC | 00:11 | |
| *** rpierce99 has joined #xdandroid | 00:58 | |
| *** rpierce99 has quit IRC | 02:10 | |
| *** kalemas has joined #xdandroid | 02:15 | |
| *** bzo has quit IRC | 02:17 | |
| *** kalemas has left #xdandroid | 03:44 | |
| *** virsys has joined #xdandroid | 03:56 | |
| *** kalemas has joined #xdandroid | 04:32 | |
| *** hardwalker has quit IRC | 05:15 | |
| *** rpierce99 has joined #xdandroid | 05:28 | |
| *** rpierce99 has quit IRC | 05:34 | |
| *** babijoee has quit IRC | 05:35 | |
| *** SytheZN has joined #xdandroid | 07:25 | |
| SytheZN | I dunno what I'm missing here... | 07:26 |
|---|---|---|
| SytheZN | I'm running ubuntu 10.10 on virtualbox, 64-bit. I've just restored my clean image, re-followed the setup, fetch and build instructions but I'm still getting errors :? | 07:28 |
| SytheZN | Tried a clean reinstall as well, to no avail | 07:29 |
| SytheZN | I'm assuming its an id-10-t error again, but I don't know where I'm going wrong | 07:30 |
| *** helicopter88 has joined #xdandroid | 07:58 | |
| *** kalemas has left #xdandroid | 08:20 | |
| *** kalemas has joined #xdandroid | 08:21 | |
| *** babijoee has joined #xdandroid | 08:55 | |
| *** SytheZN has quit IRC | 09:09 | |
| *** babijoee has quit IRC | 09:14 | |
| *** kalemas has quit IRC | 09:36 | |
| *** rpierce99 has joined #xdandroid | 09:41 | |
| *** emwe has joined #xdandroid | 10:10 | |
| *** |Jeroen| has joined #xdandroid | 12:17 | |
| *** emwe_ has joined #xdandroid | 12:58 | |
| *** emwe has quit IRC | 13:01 | |
| *** |Jeroen| has quit IRC | 13:02 | |
| *** emwe_ is now known as emwe | 13:05 | |
| *** rpierce99 has quit IRC | 13:38 | |
| *** rpierce99 has joined #xdandroid | 13:39 | |
| mgross029 | Hey emwe, How goes it? | 13:40 |
| emwe | trying to get ahold of the initially so calm ringer/dialtone/notification volume. | 13:41 |
| emwe | i had already burned openSUSE 12.1 for getting this onto another partition here and testing out ... but then i could somehow motivate myself :) | 13:42 |
| emwe | dunno for how long. | 13:42 |
| mgross029 | Oh so you mean making sure that volume is handled properly when the initial alert/ringer comes in? | 13:43 |
| mgross029 | Yeah I do all of my linux on VM here. | 13:43 |
| mgross029 | Makes it easier to build and blow up. :) | 13:44 |
| emwe | yes, but also normal notifications are too calm imho | 13:44 |
| emwe | like sms or gtalk | 13:44 |
| emwe | or mail. whatever. | 13:44 |
| mgross029 | Understand. I know exactly what you are speaking of related to the sound. I was getting sms and it was garbled at times yesterday. | 13:44 |
| emwe | damnit. i always need to kexecs or killall zygotes to get service back... | 13:45 |
| emwe | i mostly miss any notification except calls because there's vibrate :) | 13:45 |
| emwe | apart from that, you got no incall mic issues i recap? | 13:45 |
| mgross029 | Also, after I said all was working well yesterday I had a kernel panic on a gps app then my ringer and mic reverted back to a problem state. | 13:46 |
| emwe | i was looking into acoustic yesterday and then asking myself.. what bugs am i chasing here... got no issues i know off :) | 13:46 |
| emwe | hm. | 13:46 |
| emwe | gps and getting a call mostly resets here | 13:46 |
| emwe | that's to be resolved on another day. | 13:46 |
| mgross029 | I've seen those resets before too. Tells me that ril and acoustics may still be tied together, which you and detule were talking about the other day. | 13:47 |
| emwe | well, there's still that mute issued to modem processor. perhaps can be dropped. | 13:48 |
| emwe | but in combination with GPS all sorts of stuff might break... | 13:48 |
| mgross029 | If anyone were to ask me what the priority of this project would be in order to get it out into public release that would be acoustic, which is sounds like that is what you are working on. :) | 13:49 |
| emwe | because we are reading out GPS data via rpc, which also the kernel drivers for the mic-bias for example. | 13:49 |
| emwe | well, i can't promise anything. | 13:49 |
| emwe | just that initial notification tone volume and i'd say just push out and be done... | 13:49 |
| emwe | we got no mass tests... nothing to judge from. | 13:50 |
| mgross029 | I noticed that when I enabled a boot up sound it seemed to help sometimes with initial notification and ringtones. | 13:50 |
| emwe | that "sometimes" in your sentence sounds undeterministic. | 13:51 |
| mgross029 | Yeah it sounds like we are getting a few more folks back into things but until we release an updated beta we may not have enough feedback. I agree | 13:52 |
| mgross029 | Well I only tested it for a short time that way. Then I went to a fresh data.ext4 with another build and seemed to be working well after the new image | 13:53 |
| mgross029 | So, I am not sure of the consistency of my testing with a boot up sound. | 13:54 |
| emwe | bah, it's partially not picking up my pushed libs .... yet another restart... | 13:54 |
| emwe | let's see what i can find out by the logs... | 13:54 |
| emwe | it always unsets (zeros) the audio device vol when it thinks it's doing an device change. | 13:55 |
| emwe | zero volume doesn't mean off, but slightly hearable. | 13:55 |
| emwe | sounds what we got here... | 13:55 |
| emwe | anyway, back to digging. | 13:55 |
| mgross029 | Yep I remember reading that. | 13:55 |
| mgross029 | Ok. Sorry to distract. | 13:56 |
| *** Jeeva has joined #xdandroid | 13:56 | |
| emwe | nah, your welcome for a chat. been calm in here anyway ;) | 13:56 |
| emwe | my mac&cheese needs a warm microwave welcome. | 13:56 |
| mgross029 | lol | 13:56 |
| *** Jeeva has left #xdandroid | 13:56 | |
| mgross029 | Just out of curiosity could we easily build an image so the debuging is cut to a minimum and see if that helps? I noticed a little better response after you calmed the led debug on one of your latest commits | 13:58 |
| emwe | that's mostly kernel needing to drop debug output. when on wake we so much pollute the kmsg buffer it might also delay stuff | 13:59 |
| emwe | userland isn't the least verbose either ... | 13:59 |
| emwe | i dunno if acoustic debugging is on in the last build i did. | 14:00 |
| emwe | hm, stupid. it always takes to restarts (hot restarts) in order to get network operational here... | 14:00 |
| mgross029 | Yeah that is wierd. I have issues when booting into WM. I almost always come up without network and then it causes the phone not to respond at times. | 14:01 |
| emwe | oh ;) | 14:02 |
| mgross029 | I do notice a lot of acoustic messages in logcat | 14:02 |
| emwe | not to respond in wince or android? | 14:02 |
| mgross029 | wince | 14:02 |
| emwe | are they prefixed with V/? | 14:02 |
| emwe | like this? 11-25 19:57:22.371 V/AudioHardwareMSM72XX_wince( 7544): setP.... | 14:02 |
| mgross029 | IIRC yes | 14:02 |
| emwe | then yes, verbose debugging was on then. | 14:03 |
| mgross029 | Correct | 14:03 |
| mgross029 | Those are the printk statements correct? | 14:03 |
| emwe | nope, logcat. | 14:05 |
| emwe | printk is dmesg/kmsg in kernel | 14:05 |
| emwe | userland has logcat. | 14:05 |
| mgross029 | Ah ok. | 14:06 |
| mgross029 | Thanks for the clarification | 14:06 |
| emwe | man, this doesn't really sound like a libaudio/kernel issue with those sounds. | 14:08 |
| emwe | because all the volume/level adjustments are computed by audiomanager. | 14:08 |
| mgross029 | audmgr.c right? | 14:08 |
| emwe | what bugs me is that qualcomms librarry has a volume range from 0-7 .. we only 5. | 14:08 |
| emwe | nope, the kernel land one. | 14:09 |
| mgross029 | Hmm that is interesting. | 14:09 |
| emwe | and when i slide up, it get's ultra loud and back to the old position it is louder then before | 14:09 |
| mgross029 | I noticed in the logs that you saw we were splitting some things out unlike Code Aurora | 14:09 |
| emwe | then after 1-2 secs audio devices are shut off again. | 14:09 |
| emwe | (visible in kernel) | 14:09 |
| emwe | and then you start over again.... | 14:10 |
| emwe | so as soon as audio devices are turned off, it's all reset. | 14:10 |
| *** ImCoKeMaN has quit IRC | 14:10 | |
| emwe | but if you don't let them go off after some idle time the "new/louder" volume will persist. | 14:10 |
| emwe | going looki looki :) | 14:11 |
| mgross029 | Which would solve wake up with no audio right | 14:11 |
| emwe | pardon? | 14:11 |
| emwe | i am just investigating the "sound is too calm initially" stuff | 14:11 |
| emwe | do you mean that with wake up with no audio? | 14:12 |
| emwe | too calm ringer vol? | 14:12 |
| mgross029 | Yes | 14:12 |
| mgross029 | Or even notification volume is too calm | 14:12 |
| emwe | yeah ok. | 14:13 |
| emwe | so, userland volume bar has 7 (!) steps. that's it, i am sure. | 14:13 |
| emwe | we only support 5 as far is i grasped the whole years. | 14:14 |
| emwe | likely our acoustic csv do the same. | 14:14 |
| emwe | now let me figure where we can tell userland we only support 5 nuances. | 14:14 |
| mgross029 | I would agree. That is all I see is 5 in the code | 14:14 |
| *** mgross029 has quit IRC | 14:15 | |
| *** mgross has joined #xdandroid | 14:16 | |
| *** mgross is now known as mgross029 | 14:17 | |
| mgross029 | Sorry server rebooted on my. -_- | 14:18 |
| *** mgross029 has quit IRC | 14:22 | |
| *** mgross029 has joined #xdandroid | 14:23 | |
| rpierce99 | at least on mine, i can hear the difference between all 7 levels | 14:23 |
| mgross029 | rpierce99, how do you determine that you are hearing those additional 2 levels? Just curious | 14:25 |
| rpierce99 | i just hit vol up/down 7 times and listen to each, they are noticably different | 14:26 |
| rpierce99 | just the little "beep" when changing volume | 14:26 |
| mgross029 | Yeah you are right. | 14:26 |
| mgross029 | I notice that here too. Never realized that before. Seems to almost go a bit higher. | 14:27 |
| mgross029 | Actually if I keep pushing the up volume it just keeps getting louder. | 14:28 |
| rpierce99 | agreed, it is capable of getting very loud but the volume slider runs out of room maybe midpoint | 14:29 |
| mgross029 | I noticed I really don't hear the beep any longer when I turn it down to mid volume | 14:29 |
| emwe | go to tone volume settings ringtone .. change slider to 50% and remmeber/listen how loud it is, then slide to max and then back to 50%. it will be louder on 50% then before. | 14:29 |
| mgross029 | ok | 14:30 |
| rpierce99 | if i put my ear up right next to the speak i can still make out a beep all the way to the lowest setting, but its a level that is really not practical for a ring | 14:31 |
| rpierce99 | s/speak/speaker/ | 14:31 |
| emwe | but these lower levels are "leveled" wrong in software /methinks | 14:32 |
| emwe | let me go play a bit... | 14:32 |
| mgross029 | Interesting to note there are 7 levels on the volume slider | 14:32 |
| emwe | that's what i said like half an hour ago. | 14:34 |
| emwe | :P | 14:34 |
| mgross029 | Sorry didn't not even catch that. lol | 14:35 |
| mgross029 | I even just reread it in the logs... lol | 14:35 |
| emwe | perhaps i am chasing a placebo here. I'll see. | 14:35 |
| emwe | haha... or we haven't updated libaudio correctly because it was made for froyo... lol | 14:36 |
| emwe | just looked qualcomms initial gingerbread port of their froyo code... | 14:37 |
| emwe | - int vol = ceil(v * 5.0); | 14:37 |
| emwe | + int vol = ceil(v * 7.0); | 14:37 |
| emwe | good morning. | 14:37 |
| mgross029 | Wow that could be an issue | 14:37 |
| rpierce99 | well i think that implies that they knowingly changed it, so 7 is likely correct | 14:38 |
| emwe | somebody got a froyo setup around running? | 14:39 |
| rpierce99 | not running but i can easily reboot into froyo | 14:39 |
| mgross029 | rpierce99 beat me to it. :p | 14:40 |
| emwe | don't waste your precious time guys | 14:41 |
| mgross029 | http://gitorious.org/liquide/hardware_msm7k/commit/d437fe26e915924f2cc1a44ffa0bc8bd3323543a/diffs | 14:41 |
| mgross029 | Interesting. Looks like CA did it when they went from eclair to froyo | 14:42 |
| mgross029 | Made the change from 5 to 7 that is. | 14:42 |
| emwe | hm, why did this issue never crop up in froyo then for us? | 14:44 |
| emwe | or is it around there too? | 14:44 |
| mgross029 | Well we had some audio problems with froyo. Acoustics always seemed to be a problem | 14:44 |
| mgross029 | Static in speaker, etc, etc... | 14:45 |
| emwe | and this specific issue? | 14:45 |
| mgross029 | Not sure its been so long since I've used froyo | 14:46 |
| mgross029 | Checking AudioHardware.cpp I can't even find it in ours. | 14:48 |
| mgross029 | Nevermind. User errors. :p | 14:49 |
| emwe | we got other places though where we skip vol>5 | 14:49 |
| *** kalemas has joined #xdandroid | 15:10 | |
| *** kalemas has left #xdandroid | 15:10 | |
| mgross029 | Found this entry volume = ceil(volume * 7.0); as well at lrint(v * 5.0) | 15:22 |
| mgross029 | s/as well at/as well as | 15:23 |
| *** jonpry is now known as invite | 15:26 | |
| *** invite is now known as jonpry | 15:26 | |
| emwe | hm, damnit. there a "volume * 6" piece of code for the tpa2016d2 speaker amplifier... which looks bogus, too. | 16:00 |
| emwe | the tpa2016d2 supports -28dB to 30dB, so 59dB steps. then volume * 8 sounds more logical... given the register works like that ;) | 16:03 |
| mgross029 | As the rabbit hole gets deeper. :) | 16:06 |
| emwe | i see qualcomm pushing that userland volume range (0-7) just off to kernel which then communicates that via rpc call. now what if our CSVs and the acoustic data we upload don't even support 7, or rather 8 settings... | 16:08 |
| emwe | i should look at those csv, eh? ... | 16:08 |
| mgross029 | Do you have them some where handy? If not I can get them easily | 16:09 |
| emwe | got them here. | 16:09 |
| mgross029 | ok | 16:09 |
| emwe | also on the sd root :) | 16:09 |
| emwe | even got yours here with the xda mods | 16:11 |
| mgross029 | Looks like 0-5 in AudioPara3.csv | 16:12 |
| emwe | yappediyapp | 16:14 |
| mgross029 | So I wonder what will happen when I put this new system.ext2 with volume 7 in. :p heh Just experimenting over here. | 16:17 |
| emwe | let me tell you i fooled myself? | 16:18 |
| emwe | userland will report set from 0.0 -> 1.0 ... so there is nothing stopping us from multiypling this by 5 ... | 16:19 |
| mgross029 | by 5 or 7? | 16:20 |
| emwe | perhaps they updated their csv with 8 (0-7) settings for their devices. | 16:20 |
| emwe | by 5 | 16:20 |
| mgross029 | Ok | 16:20 |
| *** detule has joined #xdandroid | 16:34 | |
| *** helicopter88 has quit IRC | 16:37 | |
| emwe | ok, the tpa2016d2 gain register definitively operates in dB. | 17:01 |
| mgross029 | ok so does that mean it will needed stepped up? | 17:04 |
| mgross029 | s/needed/need | 17:06 |
| emwe | the volume level (0-5) needs to be transformed into the possible dB range. | 17:06 |
| emwe | 0 register value is actually 0dB ... | 17:06 |
| mgross029 | Ok. | 17:07 |
| emwe | one can set from -28 to +30, but only if compression ratio is off... just checking what we set | 17:07 |
| emwe | otherwise if compression ratio is on, only from 0-30dB | 17:07 |
| mgross029 | I see. | 17:07 |
| emwe | ok, compression ratio seems off, so we can distribute the 0-5 among all possible values it seems. | 17:08 |
| mgross029 | So those tpa2016d2 settings take place in the kernel not userland, right? | 17:18 |
| emwe | userland sets them in libacoustic via ioctl to the tpa2016d2 driver, which in turn sends them over via i2c to the chip itself. | 17:21 |
| emwe | s/sets/sends | 17:21 |
| mgross029 | Ok. Understood. | 17:21 |
| mgross029 | emwe, going to get away from this for a bit. Laters | 18:17 |
| emwe | yeah, will likely postpone till tomorrow as well | 18:19 |
| emwe | laters mgross029. | 18:19 |
| mgross029 | Yep getting pretty late there. Tomorrow's another day. :) | 18:20 |
| *** babijoee has joined #xdandroid | 19:00 | |
| *** detule has quit IRC | 19:13 | |
| *** detule has joined #xdandroid | 19:17 | |
| *** detule has quit IRC | 19:37 | |
| *** detule has joined #xdandroid | 19:48 | |
| *** rpierce99 has quit IRC | 19:53 | |
| *** detule has quit IRC | 19:56 | |
| *** emwe has quit IRC | 20:31 | |
| *** babijoee has quit IRC | 21:08 | |
| *** rpierce99 has joined #xdandroid | 21:43 | |
| *** babijoee has joined #xdandroid | 22:03 | |
| *** [7] has quit IRC | 22:29 | |
| *** TheSeven has joined #xdandroid | 22:29 | |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!