[13:21] <demhlyr> and it takes several dozen to hundreds of attempts [01:21] <b0000n> hello, a question: has someone used a RTL2838 dtv stick to tune isdb-t channels? [01:27] <b0000n> when i use $ dvbv5-scan ar-Argentina [01:27] <b0000n> it fails [18:12] <mchehab> b0000n: most rtl283x devices don't really have support for ISDB-T [18:13] <mchehab> although some are sold as if they would support 1-Seg [18:13] <mchehab> those devices actually do ISDB-T decoding in software [18:13] <b0000n> ok [18:14] <mchehab> currently, I don't know any open source software decoder for ISDB-T [18:14] <b0000n> i was asking because in windows it can decode a channel where i am [18:18] <b0000n_> i was saying [18:18] *** b0000n has quit IRC (Quit: b0000n) [18:18] *** b0000n_ changes his alias to b0000n [18:19] <b0000n> because in windows it is possible [18:21] <b0000n> at least with the dongle that i have [18:22] <b0000n> and i buy it as a dvb-t in aliexpress [18:22] <b0000n> (also are the logs of the channel not working?) [18:26] <mchehab> b0000n: the driver comes with a software decoder on windows [18:26] <mchehab> the device sends samples to the software decoder, using SDR mode [18:27] <mchehab> you can use your dongle on SDR mode on Linux too [18:27] <b0000n> in windows? [18:27] <mchehab> but you'll need a software decoder [18:27] <mchehab> in userspace [18:27] <b0000n> yeah i bought it for sdr and because it was cheap [18:27] <mchehab> yes, it is cheap because it is crap ;) [18:27] <mchehab> (at least for isdb-t) [18:28] <mchehab> it is just 1-seg (as the SDR doesn't provide enough samples to decode the full 6MHz spectrum) [18:29] <mchehab> and probably requires a lot of CPU for decoding ISDB-T [18:29] <b0000n> a yeah [18:29] <awalls> SDR always requires a lot of CPU to do anything useful. [18:29] <b0000n> in windows doesnt seem to suck much cpu [18:31] <devinheitmueller> awalls: hey! Long time no see! [18:31] <b0000n> so in short it isnt gonna have isdb-t support in the mean time [18:31] <awalls> Hey devinheitmueller. [18:32] <devinheitmueller> Nobody is going to bother writing a userland decoder while hardware solutions are so cheap. It's the same thing as "Winmodems" back in the 1990's.' [18:32] <awalls> http://lists.gnu.org/archive/html/discuss-gnuradio/2013-12/msg00462.html [18:32] <b0000n> seems rasonable [18:32] <devinheitmueller> awalls: Sorry, should have been more clear - I mean aside from GNU radio. [18:33] <awalls> Not even a good implementation for GNURadio yet. [18:33] <devinheitmueller> And of course GNU radio isn't going to expose the decoded stream in a way that the standard applications will supporrt. [18:33] <awalls> University and research type stuff. [18:33] <b0000n> and how it is possible to support dvb-t? it does it in hardware? [18:34] <devinheitmueller> awalls: CrestaTech had a pretty nice solution from what I hear. [18:34] <awalls> b0000n: DVB-T is demodulated in hardware and an MPEG-TS is output by the chip, usually. [18:34] <mchehab> b0000n: yes, it does dvb-t in hardware [18:35] <mchehab> the dvb-t decoder is integrated at this realtek chipset [18:35] * awalls googles the CrestaTech [18:35] <devinheitmueller> s/crestatech/xceive/ [18:35] <awalls> Ah. [18:36] <awalls> SDR tuners from the old SDR tuner people [18:36] <devinheitmueller> Speaking of demods, have you seen what silabs has been up to lately? [18:36] <devinheitmueller> http://www.silabs.com/products/video/demodulator/Pages/si218x.aspx [18:36] <devinheitmueller> Single chip: ISDB-T and DVB-T2/C2/S2X/S2/T/C/S Digital Demodulator* [18:37] <awalls> Ooh viterbi on chip. [18:38] <devinheitmueller> Although not in the headline, it also does QAM-B (i.e. ClearQAM for US) [18:38] <devinheitmueller> No 8vsb though. [18:38] <awalls> Hmmm. They have hardware demods followed by DSP. [18:39] <awalls> So yeah, they'd have to add an 8-VSB demod to the chip [18:39] <awalls> Err wait, maybe I just don't understand the block diagram [18:39] <devinheitmueller> Pretty nice though. If it had 8vsb then it would take care of 99% of my customer use cases with a single chip. [18:40] <awalls> Yeah it looks nice. [18:41] <devinheitmueller> They have a whole series of dual demods two, essentially at the cost of going from a 7x7mm footprint to 8x8mm. [18:41] <awalls> I've been doing a lot of high throughput SDR stuff lately. (Chew up 18 cores on a 32 core x86_64 machine) [18:41] <devinheitmueller> Nice. [18:42] <devinheitmueller> With a system that fast it could probably compile the whole Linux kernel in less than 30 minutes. :-) [18:42] <awalls> It's nice, but the fans are noisy. :P [18:42] <devinheitmueller> I would imagine. [18:43] * awalls needs to get back to his annual report contract deliverable instead of stalling in IRC :) [18:43] <awalls> ttyl [18:43] <devinheitmueller> I did a boot loader project some years back on such a platform (where I needed to be able to physically whack the switch). I think it caused hearing loss. [18:43] <devinheitmueller> Ok, ttyl. [18:44] <awalls> b0000n: https://github.com/git-artes/gr-isdbt [18:45] <b0000n> thanks will try to compile later [18:59] *** tuxuser changes his alias to fluxuser [19:15] <b0000n> another thing... how i can see in windows if the hardware is spiting the raw signals it get or if it is spiting a mpeg stream? [19:16] <devinheitmueller> USB analyzer? [19:17] <devinheitmueller> https://wiki.wireshark.org/CaptureSetup/USB#Windows [19:18] <devinheitmueller> I've used this one before: http://www.pcausa.com/Utilities/UsbSnoop/ [19:18] <b0000n> nice [19:19] <devinheitmueller> If money isn't a concern, these work great: http://www.totalphase.com/products/beagle-usb480/ [19:44] <joao> hola [19:45] *** joao has quit IRC (Quit: joao) [19:56] <b0000n> money is a concern. thanks for the help [20:11] *** b0000n has quit IRC (Quit: b0000n)