Root Access

I even installed Samsung drive manager

Wow.

theguardian.com/world/2014/jul/09/…

but still can't see Samsung external hard drive

;-)

Let me guess what is wrong here:
An aircraft flies at **low** altitude over a bird reserve.
It then runs into birds and crashes.

7:16 PM

@Boris_yo There is a Superuser for ID conflicting disks here superuser.com/questions/382244/…

@Bob nice :D

@Bob finally got gateone working over https with my startssl cert, under apache...

@ÃŁŁǫǛȉЖΦΤїҪ Hahah yes! Also, just got up to the part where they found Earth.

@Hennes Did you guess that after reading the article? Or just by looking at the title? Because that's exactly what happened. :(

The birdwatchers on the ground were a fairly decent hint

@Hennes You forgot in the dark, with night vision goggles. It is ok though they will carpet bomb the reserve after a detailed intelligence assesment and reduce the threat to all americans (dum enough to do the same thing).

Yea, the heli was flying at 200km/h only 30m from the ground near a reserve. The noise spooked the flock, which took of straight into the heli.

7:29 PM

But people dying that way improve the race!

the only option is to do the restart i think

but i remember that it was also possible to somehow trigger the hard drive in control panel

i think it was possible in windows xp

i mean the hard drive that was hidden

to kind of wake it up...

;-)

i think i have been doing it with wifi network card

or usb memory stick?

i don't remember now

@JANORTS why would you need a driver for a simple usb connecting hard drive ?

i dont know i am thinking why it is not showing up

do i have to do the restart? then it will show up among the hard drives in Explorer?

Some of the usb connected drives can go to sleep, and do a dissapearing act. You insure that it does not need safe removal, then pull the power, and put it back in.

it should work easily - > i plug in the hard drive to USB port and it should show up in Explorer

7:36 PM

Yes it should. there has been 2-3 questions on SU about that not happening the way it should.

I forget if anyone claimed to have known the reasons why.

and the solution?

you dont remember either?

;-)

when i unplug it from power and plug it in again... the white light on the hard drive is not lit

so it means that now it is sleeping

If i knew a direct solution, i would tell. I think they "solved it themseves" which might suggest simple restarts, checking for port problems, or (aggg) returning the hardware. or something they would tell no one, they had not plugged in the adapter :-)

so i have to unplug the cable from USB port , plug it in... and then the light will be lit

but still, it wont show up in Explorer...

it used to work 2 months ago...

i didnt use it for 2 months though

@JANORTS then go back further. see if the disk controller shows up in device manager first. then disk manager for the presence of the drive.

when i unplug it from USB port....and plug it again.... the light is lit... and also there are some strange sounds coming from the hard drive

7:42 PM

Yea you should know how your white light was acting, (shows actual data transfers) and they were stupid to not have put other leds on the controller.

in Device Manager I only see the internal HDD in my laptop

nothing else

under Disk drives

it looks like it is not communicating with Windows 7

like if the firewall was not letting it to be connected to Windows

or like if the driver was missing

I cant make it connect to guest OS right?

in vmWare Player

when it is not connected to host OS... it cant be connected to guest OS

right?

I think i will try to restart the laptop and maybe it will work then

;-)

@JANORTS In device mananger you can see everything, that ever was, but you have to set to view hidden devices, and it also helps to tweak the enviroment variable to show all devices. This does not help to get a device working, it only helps to see everything .

ok, i clicked on SHOW HIDDN DEVICES

but still nothing

it does not show up in Disk drives

Device has to exist first, so that is the lowest level.

If you ever want to see everything in a windows device manager, you can do this tweak.

It does not "fix" anything, but it allows to observe everything added with the View-Show Hidden. Beats putting another driver on something that is already drivered and in wait for enumerations.

8:05 PM

where will i find system properties?

control panel and then where?

system?

ok

in win7 it is one extra step, in the haystack where the needles are found :-)

ok

and i have to create that variable right?

"Advanced System Settings"

Yes the variable does not exist in default.

8:08 PM

DEVMGR_SHOW_NONPRESENT_DEVICES

you can find this tweak all over the web, i aint about to try and spell it :-) but copy paste works.

;O)

and now?

it should show up in device manager right?

;-)

it is not there

;o)

Then you have to reboot, then you have to turn on Menu item View-Show Hidden Devices in the device manager. and then you get a huge pile of shit you never saw before, and your eyes glaze over for a while.

ok

i will do that in 2 hours

after ARGENTINA - HOLLAND match

is over

;o)

44 secs ago, by JAN ORTS

;o)

8:11 PM

or maybe during half time

@Bob @OliverSalzburg This guy looks familiar...

Oliver Salzburg

@ÃŁŁǫǛȉЖΦΤїҪ Should we have him killed?

@OliverSalzburg if his body can make many thousands of Pringles® potato chips, sure!

;o))))))))))))))))))))))))))))))))))))))))))))))))

there is no way to prove that I am who you think I am

;-)

@JANORTS I know who you are. you are EinsteinsGrandson!

8:17 PM

Voldemort

no one else writes half-sentences on each chat line and uses ;o) after every statement

;-)

there can 2 people like that on the earth ... out of 8 billions

what if i donate 100 USD each year to StackExchange? then i can do whatever i want here? ;-)

or should it be 500 USD?

@JANORTS StackExchange doesn't take donations; they are a for-profit company, not a charity

but i can donate to a for-profit company anyway , can't i?

$500 would only pay for 1 moderator for one day :-)

8:20 PM

haha, ok so 500 USD for each day i use SuperUser chat?

;-)

(but dont tell the mods that arent getting anything:-)

how much money do i need to buy StackExchange?

1 billion USD?

4.8 million

that's not that much...

Oliver is not here but tell him I will invite him for a beer next time i am in Frankfurt, if he didn't move from there yet

;-)

he hates from the bottom of his heart, i want to meet him in person

@JANORTS sure for $500 we will let you use the chat for 1 week, let me get my paypal out so you can send that more directaly Psycogeek@nigerianscam.com

8:26 PM

1 week is not enough.... i need it for 1 year

Hello

@Hennes Remember when I was last here?

I had a faulty screen...

I've bought an Amazon replacement

But the new screen is just [BLANK]!!!!

@JANORTS not to worry it is a recurring subscrition, just keep filling up the payPal and we will keep emptying it for you.

i.e. working, but white

And now that I've seen this: ifixit.com/Answers/View/46874/…

I really don't want to find out that I have to send back my new screen

@Everyone So could there be a reason why it is just blank?

Thank you

@TimTimmy So does a LCD light valve remain open or closed when there is no charge going to it? Open , so a completly white looking screen would susgest that the light valves are not being activated. Cool thing though is, the backlight is working :-)

Ah

I just realised that the model number I asked isn't the same as the one I received!

8:37 PM

@Psycogeek valves? oO

Tut tut Amazon

So what now?!

You contact Amazon.

@Psycogeek I'm sorry, I only understand half of your message

There's some strange grammar in it.

@MichaelFrank Amazon is such a pain! :(

So is trying to troubleshoot a DOA monitor over the internet.

Light valve

A light valve (LV) is a device for varying the quantity of light, from a source, which reaches a target. Examples of targets are computer screen surfaces, or a wall screen in the case of a light projector. There are two basic principles of achieving this. One is by deflecting the light on its way to the target (a reflective LV). The other method is to block the light (a transmissive LV). The blocking method has found its way into liquid crystal flat screens (LCDs), video projectors and rear projection TVs. In this type of screens and projectors, the source light is first polarised by a f...

8:39 PM

@Psycogeek ...TIL

Anyone ever recover a corrupted PDF?

ummm

I'm really wary of what Google turns up, seems like a spammy pile of junk...

Well I got some software for uncorrupting pictures once, and it didn't work

But have a look at this one: download.cnet.com/SysInfoTools-PDF-Repair-Tool/…

@TimTimmy Corrupted is corrupted - pretty hard to right that ship

8:43 PM

4 stars - that should be reliable?

@TimTimmy CNet is buggy

But maybe it might work

well well, you might at least try

@TimTimmy only five votes

Agg Cnet is malware fest now.

@studiohack corrupted how? truncated? random data written all over it?

8:45 PM

@ÃŁŁǫǛȉЖΦΤїҪ I wouldn't know - just that Acrobat or Foxit won't open it, says corrupted or "Not a PDF file"

@ÃŁŁǫǛȉЖΦΤїҪ nice lettering! :)

@TimTimmy I remember, but I have not much experience with this.

@studiohack I would fire up a "hex editor" of some sort and look at the text (not the hex) of the header or that can also be done in a notepad. See if there are any clues in the matrix. Like if it was rared or zipped or some such oddity, there might be a clue in the header.

@Psycogeek I did that with Notepad++ but I can't figure out how to look at it

all that meaningless code is hard to look at

Could be a clue there in the header

Or in the end, to see if it is "complete"

@Psycogeek I look for "complete"?

8:58 PM

It is the matrix, only the agents know what to look for, for sure.

(end stream , enddob)

To know more you take a class in forensic Re-downloading , and download source discovery :-)

Oliver Salzburg

Just try to find some PDF parsing, open source project, attach a debugger, load the PDF and see where the problem is. It'll be fun!

yay! The screen works'

After some fiddling around with the wires, I got it to work

Unfortunately, for some reason, what is visible on the screen doesn't extend all the way across the screen

What may be the reason for this?

Is it because it has a different resolution? I have absolutely no idea

Thank you

@TimTimmy sure could be wrong res and still"work", or the display needs to be re-initilised.

by any chance , don't you know why is my Java not working for Online Banking?

it works in Ubuntu but does not work in Windows 7

bank is using an user interface that is made in Java

when fill in login and password it just freezes

@Psycogeek Do you see what I mean?

^^Like that

@Psycogeek Do you mean re-initialised?

I guess so.

9:12 PM

@TimTimmy been there done that, you said it was not the correct model, so get "data" of the 2 items, test your theory that the res is different on the pannels. or Fully reset the device, which would reinitilise the display, but most of these mobile things have only one res. Some have 2 Different connection points for different pannels (like a china thing i had once).

@Bob Wow, I'm converting one of that guy's videos on my server using libvpx, libopus and ffmpeg from git, into WebM/{VP9,Opus} format, with a total bitrate (between the video and audio) of about 19 to 20 kbps... gonna see how awful it looks/sounds :P it might end up being actually ok, surprisingly O_O

postimg.org/image/4fao5ezmh

@Psycogeek brb 10mins

you can tell the computational toll of more advanced algorithms though: I was getting about 330 fps out of encoding into libx264 and aac, but I'm getting 29 fps with vp9 and opus

@ÃŁŁǫǛȉЖΦΤїҪ Interesting choice of name there!

9:21 PM

8% muxing overhead :D

@Psycogeek Je vois, mais qu'est-ce que vous me proposer de faire ?

Aussi merci pour vos messages

@studiohack you did see me mention trying to unzip a faulty downloaded file? In rare cases a person will have compressed the file with some method prior to upload, and the extention is incorrect for the file. 7-zip (for example) would try to unzip many different compression methods by getting info from the header. So it would ignore eronious .PDF if it was actually a .RAR or .ZIP, or something. Once un-compressed, it might be a real file of some sort.

@TimTimmy I suggest you realise that I cannot even type english very well, so typing in some other language is not going to help, me being native english.

@TimTimmy Get the correct pannel for replacement. or learn how to touch things completly offset :-)

9:38 PM

@Bob from 5.5 MB down to about 3.4 MB with really excellent quality for vp9 and libopus O_O

tiyukquellmalz.org/awal/orig/encoding_command.txt

@Psycogeek thanks a lot for your time!

Bye

I'll follow your advice

@slhck You may be interested in our futzing around with ffmpeg :D tiyukquellmalz.org/awal/orig/encoding_command.txt

We started out with a bunch of coursera lectures encoded using Quicktime at half-HD res with 128k AAC audio and H264 level 3.1 video. Since the lectures have a ridiculous amount of solid white color in the background, a Constant Rate Factor of 50 (out of 68) in libvpx-vp9 gives excellent quality text and handwriting on the slides. libopus can do the voice at listenable quality at 8 kilobits per second.

The nature of content is extremely high-contrast video with a lot of solid colors, infrequently changing frames, and the audio is 100% voice (a guy talking). It turns out that you can succinctly store this information in a way that is psychoacoustically and psychovisually acceptable in about 1/8th the bits used by the already-efficient Quicktime encoder. O_o

28 MB down to around 3.5.

and I mean, H264 is nothing to shake a stick at -- we're not talking about source data that's zipped DV data, or Microsoft MPEG-1 or any kind of crap like that

AKA Talking Heads is easily compressable because nothing much changes.

@Psycogeek right, but even the "advanced" H264 is nothing compared to how well VP9 can do it :)

sure and as soon as they get VP9 into all my mobile and non-upgradable devices i will then care :-)

Until then it is just another standard added to the other 14

9:49 PM

@Psycogeek VP9 is there on my phone... Chrome and FF

wonder if MXplayer has that covered well yet? meaning it doesnt puke out everyonce in a while.

it's there on my desktop... FF and Chrome

dunno about mxplayer (I have that but I didn't listen to the video using that) but it works fine on firefox and chrome on android

For me it is offline that matters, because not everyone gets to use thier grandfathers :-) mobile data payment plan

There have been other "tricks" in video codecs that were not applied very well, and not supported fully by all the decoding/players, that would compress further without quality losses. The New codec will be able to apply every trick in the book, and all the decoders will actually work right, when using all the tricks.

10:26 PM

@Bob @Awal 10.webm == 1.7 MB, 10.mp4 (original) == 12 MB... and it looks and sounds great!

even on Android :D

There is Media player classic ^ actually playing a standard resolution video , the ammount it takes of my Quad

Plus the audio there taking ~2x as much effort ? plus 10% of the gpu clocked at 1/4 rate

When doing offline mobile with the greatly improved codec, and grafics acceleration and all, I would want to see some of the Power stats. there might be a balance between the storage data (power to move data) and CPU effort to playback. How much battery life each takes for exstended media playing while un-plugged.

@Psycogeek smaller media involves less memory and storage I/O, so with GPU accelerated decoding, smaller media takes less power to play

Hardware accelerated GPU processing takes way more power.

as codecs have gotten more complicated, encoding time has exploded very very fast... however, decoding time has remained small, and the rate of growth of decoding time (complexity) has been very shallow

@Psycogeek no, it doesn't

a GPU can decode a video much more power-efficiently than the CPU

the GPU's transistors are laid out to perform many independent computations in parallel; the CPU's transistors are laid out to perform many dependent computations in serial

10:34 PM

@ÃŁŁǫǛȉЖΦΤїҪ Could, but doesnt. there has been vast effort on desktops to get the cpu to take like zero power when doing zero work, the GPU is still a power vaccume.

most operations on videos are embarrassingly parallel; you can wake up the GPU, decode many seconds of video at once in a very short time (divide the video up into tiny squares), then go to sleep... while the CPU would still be working on it

@Psycogeek That's simply incorrect

I tested my mobile device with both "software render" and Hardware accelerated renders, monitoring the miliamper consumption, and there was not a visable difference in the power consumption over hours of playback. Of course i only am testing to see (about) how long my battery would last, and more real longer term and lab type testing would be useful.

@ÃŁŁǫǛȉЖΦΤїҪ If you have a meter you could put on your desktop that reads wall power, you could test it. GPU rendering is a great feature, nobody ever said it would use less power.

@Psycogeek desktops are designed completely differently than mobile devices -- if a desktop GPU isn't very efficient, very few people give a crap... they care about performance, noise, that kinda thing

on devices that are battery-powered, the CPU will be awake for longer than the GPU to decode the video, and the CPU has about the same or sometimes more TDP than the mobile GPU in such devices

of course it also depends on how you're doing the GPU decoding

You have android? check out CoolTool, where you can put up the power consumption on a overlay, and have it constantally visable , see for yourself.

if it's fixed function, it's going to be more efficient; if it's OpenCL, it'll suck

10:45 PM

I set-up my android to basically stock always display the power consumption, just to have a total idea of where it all goes (and comes from when charging). But i have one of them LCD screens not LED, so much and much of the power when mobile outdoors is the backlight.

@ÃŁŁǫǛȉЖΦΤїҪ And the effort required to decode, which is what would be an interesting stat.

@allq could you paste the ffprobe results? I'm on mobile.

@ÃŁŁǫǛȉЖΦΤїҪ ^

though, if that's acceptable voice quality then I can push aac down to 10, maybe even lower

I stopped at 16 because it introduced noticeable distortion (the tinny effect), but still perfectly understandable for voice

the video is definitely better though. I wonder what the video bitrate is. 264 at 40crf produced about 8-10kbps I think

@Bob I consider it acceptable voice quality, personally, and there's no tinny effect at 8 kbps (kilobits, not kilobytes) of libopus.

total bitrate was hovering around 24 with 16 cbr audio

10:52 PM

I also save a tremendous amount of data by cutting off (zeroing out) everything about 8 KHz frequency -- so all the background sounds of air, etc. get chopped

the PSTN is gated at around there, IIRC. maybe even lower.

@allq by the time you go that low there's far more distortion than just tinniness

@ÃŁŁǫǛȉЖΦΤїҪ ^

similar on acc

aac

> ffprobe ../tiny/10.webm
ffprobe version N-64559-ge932ae5 Copyright (c) 2007-2014 the FFmpeg developers
built on Jul 9 2014 20:54:39 with gcc 4.7 (Debian 4.7.3-4)
configuration: --enable-gpl --enable-nonfree --enable-libvpx --enable-libopus --enable-shared --enable-avresample --prefix=/usr --libdir=/usr/lib/x86_64-linux-gnu
WARNING: library configuration mismatch
swscale configuration: --prefix=/usr --extra-cflags='-g -O2 -fstack-protector --param=ssp-buffer-size=4 -Wformat -Werror=format-security ' --extra-ldflags='-Wl,-z,relro' --cc='ccache cc'

Mpc in a software render, same video . of course the CPU has all them special co processors and instruction sets too, most of which are used for video stuff.

10:54 PM

also, the aac encoder defaults to a 14khz cutoff

13.webm -- my processed version is 14.8% the size of the original

@Psycogeek software rendering and software decoding are not the same thing

it could still be doing hardware decoding under the hood

@ÃŁŁǫǛȉЖΦΤїҪ want me to remove the gpu :-)

Yes indeedy there has to be a gpu of some sort to display anything.

@allq your's doesn't show stream bit rate :/

@ÃŁŁǫǛȉЖΦΤїҪ ^

I'm pretty sure I can push aac at least as low as you did opus. Not too sure about 264, but I think I'm already at a lower video rate

10:57 PM

@Bob yeah but who's going to do the ABX test to see whether aac at the same bitrate sounds as good?

@Bob if you're at a lower video rate, why are your files so much bigger than mine?

@allq double the audio rate, more muxing overhead (can't do anything about the last bit)

@ÃŁŁǫǛȉЖΦΤїҪ ^

are you using CBR mode, by the way? I'm using VBR with a target bitrate. and yeah, the low muxing overhead of webm is nice :)

as far as sounding good, as long as you can distinguish the words?

(the "pauses" in the professor's speech get virtually zeroed out)

11:09 PM

What I am basically indicating is there has been a lot of crasy assumption about how much effort it has ever taken to Playback video. what might have been a difficult task in 1978 doesnt even blip the cpu anymore, it barely takes any extra power for a mobile device to play vrses being on.

@ÃŁŁǫǛȉЖΦΤїҪ yea, CBR at 16

it's passable quality for music, I can definitely push voice lower

> VBR encoding in libfdk-aac is experimental and only works with some parameter combinations. Target a quality, rather than a specific bit rate. 1 is lowest quality and 5 is highest quality. This mode is not compatible with AAC-HE.

Also, I'm actually forced to use stereo:

> AAC-HE version 2 only handles stereo.

@Bob heh, I'm forced to use 48 kHz sample rate, because the libopus decoder in Firefox (and possibly other browsers) doesn't support anything else

@Bob VBR in ffmpeg's own aac encoder is even worse -- sounds like VBR just is not really part of AAC's design, or if it is, it's an afterthought

VBR is a key component of opus's functionality and it only really works best in VBR mode, much like Vorbis

@ÃŁŁǫǛȉЖΦΤїҪ The build-in aac encoder is considered considerably worse than libfdk_aac

and doesn't support he-aac

he-aac is the only one that works well at such low bitrates

@Bob I think the FOSS non-patent-encumbered codecs are winning, here :P HEVC is probably still better than VP9, but VP9 is better (marginally) than H264, imo.. and opus is the best codec in existence for voice data

11:26 PM

@ÃŁŁǫǛȉЖΦΤїҪ They're neck and neck at the moment, with the disadvantage of lacking (hardware) support. But they could certainly come out ahead in the future.

opus is probably worse than AAC for high entropy data (high-fi music, etc)

@ÃŁŁǫǛȉЖΦΤїҪ That's what vorbis is for :P

I wonder how it compares to speex and miles.

@Bob the really neat thing about opus that's fairly uncommon among compressed audio codecs is that its frame size is variable - you can set it high for reduced overhead in non-interactive media, or low for real-time VOIP

well, miles is pretty crap

hehe hlds101.com/voicecompression.htm

> Valve has added the support for the Speex voice encoder. By default, HLDS is set to use the old MILES codec that uses almost 32kbps of your bandwidth compared to the Speex codec that can use from 2.4kbps 15.2kbps (depending on which one you select). Be warned that Speex will use more CPU cycles than MILES... On my 500mhz server, it uses only 2 or 4 percents more CPU!

> 500mhz server

lolol

11:34 PM

speex at hlds's voicequality 5 (~15.2kbps) can actually play music decently well (IIRC it took a 16-bit mono PCM wav, I forgot the max sample rate)... pretty nice for something over a decade old

I'm going home. :3 committed my changes to the version control system we're using and updated defect status in the defect tracker we're using

finally got some work done

oh yea, @allq, @psyco, I tested that file with mxplayer, doesn't work

@ÃŁŁǫǛȉЖΦΤїҪ ^