[Freetel-codec2] Ubuntu build

An interesting exercise, although mowing the lawn is more fun... I played around with wxwidgets 3.0.0 and where it failed to build was an if-clause and the if-item was an undefined symbol. No problem, I deleted the if-clause. Build succeeded. I suspect 3.0.1 is fine too, as I found what seems to

Re: [Freetel-codec2] Ubuntu build

Well, that didn't last long. I rebooted and even with the new theme, the audio dialog box is back to gimped. Never mind... -- Open source business process management suite built on Java and Eclipse Turn processes into bus

Re: [Freetel-codec2] Stuffing fdmdv into a uC

In creating a Java version of a couple of routines in fdmdv.c, I had to do things a bit different, due to the pointers. But I did get rid of the S[] arrays out of global memory and they now go in and out of the stack. Attached is the essential bits. Not a big savings, but at least you don't have

Re: [Freetel-codec2] Stuffing fdmdv into a uC

30, 2014 at 11:13 PM, David Rowe wrote: > Thanks Steve that's a useful idea. S1 & S2 could just be locals. Hmm, > however, I need those arrays for logging/testing purpose. Let me think > about how to handle that..... > > - David > > On 01/07/14 13:37, Steve wrote: >&g

[Freetel-codec2] qpsk to bits

The original code has some comparisons to zero, but I'm wondering if that is correct. If any zero's *are* in there, they probably just need to be mapped to Q1 (which is the default). So I had this change. It doesn't seem to break anything. msb = lsb = 0; if ((d.real > 0) && (d.imag

[Freetel-codec2] Codec2 on Andoid

I don't know much about android programming, but here's a java version (it uses 8 khz sampling, so maybe needs a bit of mod). https://github.com/k5okc/jCodec2 73, Steve -- Slashdot TV. Video for Nerds.

Re: [Freetel-codec2] Codec2 on Android

oops, I wasn't thinking. The java version is a memory hog, as I used a complex class. I need to get rid of that and replace it with arrays (one day). 73, Steve -- Slashdot TV. Video for Nerds. Stuff that matters.

Re: [Freetel-codec2] CONST table size of codec2

I think the reasoning is, that floating point and memory are so cheap now, that trying to fit a design into a restricted space would just lengthen the time to profit. Why design to a fixed point $30 DSP when you can buy a $5 CPU with hardware FP. 73,Steve

[Freetel-codec2] room for speed/memory mods in _kiss_fft_guts.h

I was just looking at this file, and maybe the ongoing mods to sinf(), cosf(), powf(), floorf(), etc may help here. Also the kf_cexp() is called with a double in kiss_fft_alloc() and can be a float. Save some bytes if the ARM compiler doesn't already do it for you.

Re: [Freetel-codec2] room for speed/memory mods in _kiss_fft_guts.h

Maybe replace all the FFT's with FHT's... http://www.embedded.com/design/connectivity/4403178/Doing-Hartley-Smartly I was also thinking, since you mentioned that the kiss code ran faster than the ARM library, that maybe speed isn't the defining goal. That is, if the ARM code is fast enough, and c

Re: [Freetel-codec2] room for speed/memory mods in _kiss_fft_guts.h

I was thinking the next big thing (down the road) would be full-duplex, and then you could have two FFT's running :-) ...although maybe that might be better done with a different design (SM2000 ??)... Many years ago we used to have a TAPR 9600 baud repeater in town. It was pretty neat, as it solv

Re: [Freetel-codec2] room for speed/memory mods in _kiss_fft_guts.h

An interesting thing is you can do a fourier with a hartley base. Here's a link (fft from a fast hartley): http://extras.springer.com/2001/978-3-540-66572-4/fxt/fxtd/fht/fhtmayer.txt -- Slashdot TV. Video for Nerds. St

[Freetel-codec2] lpf_peak_pick

peak was in imaginary filter 150, so *foff would equal -106 * 1.5625 or about -166 Hz. Whereas (ix - MPILOTFFT/2) would be 150 - 128 = 22 , or about 34 Hz. I put a print statement in there, and I actually never see ix >= MPILOTFFT/2. I'm usually wrong about these things, so I'm just th

Re: [Freetel-codec2] lpf_peak_pick

be about right. Time to switch to water instead of coffee :-) Since I didn't test with a frequency offset, that's why I never saw the print statement... 73/Steve -- Slashdot TV. Video for Nerds. Stuff that mat

Re: [Freetel-codec2] lpf_peak_pick

Having worked on radars I can understand better the opening and closing Doppler velocities (+/-) but I could never remember which was which :-) Lucky I wasn't the guy pulling the trigger. Alas, my aha! is always after the send button, argh... On Wed, Aug 27, 2014 at 3:42 PM, David Rowe wrote:

[Freetel-codec2] calc_sd standard deviation

/ this.nin[0]); - It makes for a nice level control. Although I do set the sdev to 0 if not in sync... Steve -- Slashdot TV. Video for Nerds. Stuff that matters. http://tv.

Re: [Freetel-codec2] calc_sd standard deviation

Yes I did notice. I can run from 1.0 to over 10.0 on the std dev values and not much decoding changes, although the SNR does go down near the min/max values. -- Slashdot TV. Video for Nerds. Stuff that matters. http://t

Re: [Freetel-codec2] calc_sd standard deviation

Mel, that was just for the GUI display. Actually on second thought I should just let it measure (sync or not). Sometimes I over complicate things... -- Slashdot TV. Video for Nerds. Stuff that matters. http://tv.slashdo

Re: [Freetel-codec2] calc_sd standard deviation

nt digital frames before sending their encoded values to the codec (digital pre-filtering). Say, one clip has a frequency of 80 Hz, and the next has 90 Hz, maybe split the difference. Same with amplitude. Or, if the differences are too great, turn-off all the voicing bits and add comfort noise.

Re: [Freetel-codec2] calc_sd standard deviation

Expanded thought... it would add to the latency, but maybe have a filter window of say 160 ms (4 codec slices). If the two inner slices don't conform to the two outer slices, or don't represent a staircase up or down trend, then hit it with a hammer.

Re: [Freetel-codec2] Talks from DCC available

On Sat, Sep 6, 2014 at 12:33 PM, Bruce Perens wrote: > > Interesting that AMBE has to use VQ and we do not. The 1200 mode uses VQ, which is maybe why I like the sound better? But I haven't really tested it with background noise. It does make for a large table though...

Re: [Freetel-codec2] [digitalvoice] Chris' MyHDL talk materials available

When I first hired on as a tech writer in the early 80's, I was given a stack of yellow pads. I asked for a word processor computer, and they said no. So I brought in my H-89 CP/M and Wordperfect and that was that. We had "lino" Korean ladies that spoke no english, just typed what they saw, and the

Re: [Freetel-codec2] FreeDV **IN** the Flex 6000 series

I have the older 3000, and I loved the radio until my Windows machine died :-( I like the new radios because they have left the Firewire behind! Great news by the way! -- Want excitement? Manually upgrade your production

Re: [Freetel-codec2] Boat Anchor 101!

Sometimes it's cheaper just to rent, but you have to be located someplace outside of Hooterville, or Cow Town... I have a real nice O'scope that I use maybe once a year, other than that it is usually being borrowed :-) On Wednesday, September 17, 2014 6:46:51 AM UTC-5, aa777888 wrote: > > Why n

Re: [Freetel-codec2] how to play .c2 file on windows 7 64 bit ?

I think there is a c2dec.exe in there. On Sun, Sep 21, 2014 at 8:48 AM, martin.levy wrote: > good afternoon, I experimented with your codec2 on my pc via windows > binaries from here: https://svn.code.sf.net/p/freetel/code/codec2/win32/ > > and command from dos prompt: c2enc 1400 1.wav 1.c2 gwene

Re: [Freetel-codec2] how to play .c2 file on windows 7 64 bit ?

oof, you probably knew that, but I haven't used windows in awhile. The c2dec outputs a PCM raw file sampled at 8000 Hz. So if windows has a tool like audacity, you can convert the raw 16-bit samples to a WAV. In GNU we use a program called "play", and I'm sure there's some windows function/program

Re: [Freetel-codec2] SM1000 Noise and Radio Tests

The articles schematic link is old, I think this is the newest one: http://sourceforge.net/p/freetel/code/HEAD/tree/smartmic/SM1000-C/SCH-SM1000-C.pdf -- Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer Ach

Re: [Freetel-codec2] VQ work for Codec 2

rs and code books. I probably need to finally install and configure Octave. I'm not much of a Matlab programmer; however, I can appreciate its capabilities, even if just plotting. Back in my day, MathCad was about as fancy a

[Freetel-codec2] Small Diff for trigs in revision r1893

Here's a diff for getting the remaining trig functions to float size. I did a "nm" on the library and all the references to double functions were gone. FYI diff -rupN old/interp.c source/interp.c --- old/interp.c 2014-10-15 13:18:37.673035985 -0500 +++ source/interp.c 2014-10-15 12:22:14.58113594

Re: [Freetel-codec2] short-format in freedv_rx()

change it internally to +/- 1.0 maximum. Then the interface person would make whatever wrapper they need to get to that. Interestingly, the codec doesn't work at +/- 1.0 limits, and needs +/- 32767.0. I never looked farther as to why. Just went with the

[Freetel-codec2] codec2-dev 1900 svn experimental changes

This is an attempt to zero out code that is no longer used in the wild. Basically I wrapped it with __EXPERIMENTAL__ The most changes were in c2sim.c where all these tend to be exercised. The functions and codebooks aren't used by freedv. Steve diff -ruN old/c2sim.c new/c2sim.c --- old/c2

Re: [Freetel-codec2] codec2-dev 1900 svn experimental changes

I debated releasing that, because I didn't test it fully with c2sim, and mostly just with the codec2 library. Ack, good point about the generated code. I'm a netbeans IDE guy, so I always forget about cmake sometimes. I like the refactoring idea. Obviously this doesn't save a ton of memory, but

[Freetel-codec2] Small change in generate pilot baseband

I was trying to optimize this in another language (failed so far), but did notice that it could be refreshed a bit for C, as it computes sqrt(2) * sqrt(2) / 2 which can be simplified. Saves 960 * sizeof(float) - 960 * sizeof(char) = 2880 bytes The pilot_lut[] table was checked using both original

Re: [Freetel-codec2] Small change in generate pilot baseband

My lousy typing has messed me up again... Ignore previous attachment diff -ruN freetel-code-orig/codec2-dev/src/fdmdv.c freetel-code-new/codec2-dev/src/fdmdv.c --- freetel-code-orig/codec2-dev/src/fdmdv.c2014-11-17 08:50:17.912563270 -0600 +++ freetel-code-new/codec2-dev/src/fdmdv.c 2014

Re: [Freetel-codec2] Small change in generate pilot baseband

Admittedly this is just an initialization table that is completely cleared off the stack upon completion, as all this work is merely done to populate pilot_lut[] relative to the size of a variable M. Anyway, the algorithm perked my interest for some reason, so press on... plenty of megabytes left f

Re: [Freetel-codec2] FSK over FM

> Here in my neck of the woods for me to play with this I would need to have > this fit into a 10khz Channel. No problem, just set your deviation to 1.1 to 1.5 kHz. -- Download BIRT iHub F-Type - The Free Enterprise-Grade

Re: [Freetel-codec2] tx and rx bits

would be worth the rewrite, but certainly doing an AND with 1-bit everywhere will work to contain the data. Steve -- Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge

Re: [Freetel-codec2] golay23 license

Maybe negotiate a royalty, after all, it improved the modem. Maybe his kids got coal for Christmas... -- Dive into the World of Parallel Programming! The Go Parallel Website, sponsored by Intel and developed in partnership

Re: [Freetel-codec2] golay23 license

On Mon, Dec 29, 2014 at 11:18 PM, Bruce Perens wrote: > > ...I doubt the original author's remaining work is worth much more than $100. > Should I offer him that? He may just want beer. His kids may want some Colorado plant life...

Re: [Freetel-codec2] golay23 license

A quick search found a GPL golay.c and golay.h in wireshark: https://github.com/boundary/wireshark/tree/master/epan FYI -- Dive into the World of Parallel Programming! The Go Parallel Website, sponsored by Intel and devel

Re: [Freetel-codec2] golay23 license

Actually the modem code is mostly GPL (2.1) as I seem to recall when I was porting it. Probably selected the wrong license while reaching for LGPL 2.1, but there it is... On Tue, Dec 30, 2014 at 1:21 AM, Bruce Perens wrote: > We need LGPL for this one rather than GPL. The rest of the codec and >

Re: [Freetel-codec2] DSP 64x

I think you're right, the 64x is fixed, the 67x is float. But all these chips are for gold-plated projects. Very expensive. There is just a couple of routines that suck the CPU, one is the FFT. The others are the routines that call the FFT multiple times. As long as you can process 160 samples of

Re: [Freetel-codec2] Freetel-codec2 Digest, Vol 59, Issue 4

> We need to know the processing requirement of CODEC2 because as per our > requirement > we have to complete codec2 processing(encoding and decoding both) within a > given fixed > time interrupt. please provide me some authentic statistics if you have. The algorithm must process 80 PCM 125 usec

Re: [Freetel-codec2] Codec2 licensing

Is that some kind of Fedora coding? I see no reference to licenses with a "+" plus sign at GNU. As far as I can tell, the CODEC2 and FDMDV in the codec2-dev directory is LGPL2.1 and FREEDV stuff in the freedv-dev directory is GPL2.1. --

[Freetel-codec2] cohpsk

That's a pretty nice spectrum! I played with your tools to create a test bit audio, and output the spectrum. http://goo.gl/8yV64f 73, Steve -- BPM Camp - Free Virtual Workshop May 6th at 10am PDT/1PM EDT Develop

Re: [Freetel-codec2] cohpsk

Yep, I visit at least a couple times a week. It's the only new modem design I'm aware of, so it's an interesting project. -- BPM Camp - Free Virtual Workshop May 6th at 10am PDT/1PM EDT Develop your own process in accordan

Re: [Freetel-codec2] GNU Radio Bug #776

lspjnd appears to be __EXPERIMENTAL__ defined, and should not be used in production. Undefine experimental. Same with lspdt -- BPM Camp - Free Virtual Workshop May 6th at 10am PDT/1PM EDT Develop your own process in acco

Re: [Freetel-codec2] Codec 2 "stateless" between the 20/40ms blocks?

Well, it is a lossy compression :-) Wouldn't it be neat if there was a lossless compression of the same ratio? The vocoder looks at the stream of PCM audio in 10 ms segments. There is a lot of processing that goes on to encode/decode each 20 or 40 ms chunk, but I don't think any state (the voice m

Re: [Freetel-codec2] GNU Radio Bug #776

Oops, a better idea would be to add __EXPERIMENTAL__ to the codebookjnd.c file to wrap the code, and effectively map it out. Same with codebookdt.c, codebookres.c, codebookvq.c, codebookvqanssi.c. I think that's why he's even seeing the error.

Re: [Freetel-codec2] VHF Frontend for VHF experimentation

Woof, mouser has that 2N5109 transistor for $2.00 each in large quantity. I wonder if some of the components need a 15 year update, as the technology is zipping along faster than my brain can handle :-) -- BPM Camp - Free

Re: [Freetel-codec2] cohpsk

Played with the latest (modulate bits look at spectrum). The previous version was not symmetrical and looked like the filter ended too early, but this spectrum is symmetrical. Kind of spikey, but I guess that is the pilots maybe? 600 Hz BW - nice! Also looks like FDMDV_SCALE of 1000 might be too

Re: [Freetel-codec2] cohpsk

I've been porting your 650 codec. Just have the filtering part left (I think). Ah, I didn't look deep into the code, but thought maybe it was diversified in pairs. 8 more carriers would still be pretty narrow. -- One dash

Re: [Freetel-codec2] Can't seem to adjust SM1000 rig output high enough

Minor note (maybe), but it looks like you are about 40 Hz off. The center being between the two red lines. Although I think the modem works +/- 100 Hz. But the waveform looks pretty bad. Can you reduce the gain on the modem, and increase the mic gain on the rig? I'm thinking at the ends of the pot

Re: [Freetel-codec2] Can't seem to adjust SM1000 rig output high enough

I've never seen the skirts before, on the waveform view :-) -- One dashboard for servers and applications across Physical-Virtual-Cloud Widest out-of-the-box monitoring support with 50+ applications Performance metrics, s

Re: [Freetel-codec2] VHF IF

ators. Course, I don't really know if that much precision is needed anymore, with a good software tracking algorithm. Maybe FFT based, you can maybe chase some gomers drift all over the band, with CPU power of the day.

Re: [Freetel-codec2] Help

of no use for vocoding. You might as well just store the raw PCM into a file and send that via TCP/IP using an ACK/NAK protocol. Steve -- ___ Freetel-codec2 mailing list Freetel

Re: [Freetel-codec2] cross compiling FreeDV for Win32

Maybe a good time to identify the codebooks that are no longer be experimented with, and move them to the attic. -- ___ Freetel-codec2 mailing list Freetel-codec2@lists.sourcefor

[Freetel-codec2] speaker audio dropouts

Tried the latest build under Ubuntu 10.4 64 bit. There's a lot of audio dropouts. Just an FYI, haven't looked at the changes. Also the squelch no worky. ​ -- Monitor 25 network devices or servers for free with OpManager!

Re: [Freetel-codec2] speaker audio dropouts

I can't read... 14.04 On Thu, Jun 25, 2015 at 2:58 PM, Stuart Longland (VK4MSL) wrote: > > For what it's worth, Ubuntu 10.04 is no longer supported by its vendor, > and is old enough that running newer software becomes problematic. ---

Re: [Freetel-codec2] speaker audio dropouts

I just checked and I have Portaudio 19-devel. I don't remember if I added that in my playing around. Also, found that my Yaesu 857 only like 38k baud for CAT on hamlib, but that's a good speed. FYI On Thu, Jun 25, 2015 at 4:10 PM, David Rowe wrote: > Hi Steve, > > Thank

Re: [Freetel-codec2] SM1000 to radio cables

In the military (back in the 70's with analog voice scramblers ("parkhill") it required the operator to modulate their voice in a communications style, rather than say a conversation style. The example audio has the operator in what I call the conversation style. He thinks he is on the telephon

Re: [Freetel-codec2] [digitalvoice] microphones

I chose a real fine resolution on the number of FFT bins. Of course the codec will probably never see all those bins :-) But I was trying to get a feel for the mics. On Thursday, July 23, 2015 at 7:25:55 PM UTC-5, DavidRowe wrote: > > Thanks Tony and Steve, some interesting

Re: [Freetel-codec2] Uploading a custom code for STM32 on SM1000 in Linux or Windows

Look on the wiki page: SM1000 Wiki Under: "Testing, Debugging, Development Notes" it refers to the "dfu-util" Also, take a look at: STM App Note 3990

Re: [Freetel-codec2] microphone placement and codec 2

It does seem logical that you would want a slow changing pitch. A communications quality voice, rather than high fidelity (skype). If the pitch values change too much between frames, you probably get robot noises. How to enforce slow pitch changes from untrained operators. Maybe that would be a

Re: [Freetel-codec2] options wanted-codec2 stmf4 code optimization question

I think that David said he kept the kissfft only because it was faster. The kissfft however, uses twice as much memory, not being an in-place algorithm. This can be easily changed in the code though, if a slower in-place is used that still performs the task "fast enough" and you are running out of

Re: [Freetel-codec2] Problem in playing hts1a.raw

Do you have the STLINK_DEVICE environment variable defined properly to the USB port detected after the 'dmesg'. -- ___ Freetel-codec2 mailing list Freetel-codec2@lists.sourceforge

Re: [Freetel-codec2] Problem in playing hts1a.raw

Use the 'lsusb' to get the bus and device address probably. STLINK_DEVICE=001:004; export STLINK_DEVICE For BUS 1 Device 4, or whatever associates with the USB port desired. I looked at the code and if the variable is not defined it just goes along merrily until it blows up somewhere down the li

Re: [Freetel-codec2] options wanted-codec2 stmf4 code optimization question

Yes, the frame of data. kissfft uses one array for input, and another for output, but if I recall, the input is never used. Although, I'm sure that gives it a speed advantage. Anyway, it will be used on the next pass :-) I used the code from JTransforms, where the internal tables are about 2K so n

Re: [Freetel-codec2] Add branches and tags to svn?

You might consider creating an Organization. That way participants can be members. Give them their own branch. -- ___ Freetel-codec2 mailing list Freetel-codec2@lists.sourceforge.

[Freetel-codec2] Mode 700B Data Bit

then the data shouldn't be working, so I was wondering if users are receiving data OK?? Steve -- ___ Freetel-codec2 mailing list Freetel-codec2@lists.sourceforge.net https:/

Re: [Freetel-codec2] Mode 700B Data Bit

I've heard a couple of 700B users on the air. I don't think I noticed anything with their audio though, so maybe this is just little bit errors in a sea of bits, or maybe the text bit turns out to be the same as the LSP bit value 50% of the time. Another case of the human ear being highly adaptabl

Re: [Freetel-codec2] FreeDV 700(B) Fs=8000Hz patch

That is good news, as the rate conversions were like putting a Suburu engine in a Escalade (slow)... :-) -- ___ Freetel-codec2 mailing list Freetel-codec2@lists.sourceforge.net h

Re: [Freetel-codec2] Tuning Aid 700B mode

Maybe have a tune mode, where a center carrier transmits at 1500 Hz with a di-dah-di-dah-di-dah... for 700B and di-di-di-dah-di-di-di-dah for 700(whatever)... -- ___ Freetel-code

Re: [Freetel-codec2] SM1000 Menu-driven firmware

I finally found the change, way down. When you delete the file and upload a new one, I guess it doesn't go in as a diff. It says every line has changed instead of just one. -- Monitor Your Dynamic Infrastructure at Any Sca

Re: [Freetel-codec2] SM1000 Menu-driven firmware

Oh yes, I need to fire up my win-dos box more often. I keep forgetting about these particulars :-) (I fired it up last month, and it wore out my battery downloading all the updates, and I had to finally plug it in). -- Mo

[Freetel-codec2] Minor change to codec2-dev

In freedv_api_internal.h change: #ifndef __FREEDV__ to #ifndef __FREEDV_INTERNAL__ http://sourceforge.net/p/freetel/code/HEAD/tree/codec2-dev/src/freedv_api_internal.h#l38 Reason: __FREEDV__ is already used in freedv_api.h 73

Re: [Freetel-codec2] SM1000: Persistent settings now working

My neighbor had a stereo preamp that had digital controls, and it powered up from the last settings. After 10 years it only powered up from factory settings. I told him his EEPROM probably used up its number of writes. Which brings up the point to not write anything if nothing changes, as sooner o

Re: [Freetel-codec2] SM1000 and the FreeDV 700B mode

Added to the language in C99 if I remember right... No problem in GCC. On Sat, Oct 24, 2015 at 7:01 PM, Stuart Longland wrote: > > Interestingly, we seem to be passing an integer non-const variable in > the declaration of these arrays; didn't think that was allowed. -

Re: [Freetel-codec2] SM1000 and the FreeDV 700B mode

On Sat, Oct 24, 2015 at 7:55 PM, Stuart Longland wrote: > On 25/10/15 10:49, Steve wrote: >> Added to the language in C99 if I remember right... No problem in GCC. > > Ahh, figured it was a newer feature. I sometimes have to deal with K&R C. > The difference of course, i

Re: [Freetel-codec2] SM1000 and the FreeDV 700B mode

How are you doing on RAM space? On Sat, Oct 24, 2015 at 9:23 PM, David Rowe wrote: > Hi Steve, > > There's plenty of flash left for the cohpsk modem, we already have the > entire codec and fdmdv modem fitting in around half the flash. > > No need to load any program co

[Freetel-codec2] Possible bug in lpc.c inverse_filter

Problem Description: In codec2.c we have a function called codec2_encode_700b() and an instruction: inverse_filter(&c2->bpf_buf[BPF_N], bpfb, 4*N, bpf_out, BPF_N); BPF_N is 101, 4 * N = 320 (which could be M instead) The number of coeficients of the BPF filter is 101. In lpc.c we have the loca

Re: [Freetel-codec2] Tone and mixed speech with tone showing very low MOS value

I might be mistaken, but I think tones are a separate algorithm. For example, in AMBE the tones are detected and encoded into a digital word (separate from the speech) and then recreated at the other end. That is, the tones don't pass through. ---

Re: [Freetel-codec2] Tone and mixed speech with tone showing very low MOS value

Well, disregard my last, as I was only thinking of DTMF tones, not the tones you are describing. -- ___ Freetel-codec2 mailing list Freetel-codec2@lists.sourceforge.net https://li

Re: [Freetel-codec2] OT: LPC-10 License help

on it, and there are various improvements of the C output in the wild. None of these contain a license, as they are all based on federally funded research. My favorite quote, who I don't remember the guys name, was: "I wish there was a valid patent on it so that then maybe people would stop u

Re: [Freetel-codec2] OT: LPC-10 License help

heavy in PDP-10 computers. They must have had a sale. There's a snide remark about Bell Labs getting a patent in 1978. Universities didn't have the patent departments they do now :-)

[Freetel-codec2] cohpsk in java

ave plenty of memory and speed. Still though, you'd rather just have a speedy algorithm like in C. 73, Steve -- ___ Freetel-codec2 mailing list Freetel-codec2@lists.source

Re: [Freetel-codec2] Encrypting GSM Phone

Hermes was one of the first point-to-point over cellular modem ideas I believe. The algorithm doesn't look that complicated, but I would guess the difficulty is in the details. They say they could get 1200 bps through the phone, which is less than the 1600 bps used by the sm1000. As far as I know,

[Freetel-codec2] Playing with Coherent Modem

I finally found the killer bug in my coherent PSK modem version. I had to switch to complex classes though, as the C style complex arrays didn't work. Probably some type of mutable problem, and probably not a good thing with multiple threads. The problem is, my audio sounds so bad, I couldn't tell

Re: [Freetel-codec2] Playing with Coherent Modem

Interestingly, now that the modem seems to be working I could play with it a bit today, with various sub-carrier center frequencies and no diversity. I got 8.1021 dB crest factor with 7 carriers at 112 Hz spacing (Newman phases). Not much change there. With 112 Hz the spectrum used was about 900 H

Re: [Freetel-codec2] Codec+ Crypto

To answer the question, you don't really need to change the modem, but you need to add a protocol on top. Use the 1600 codec (no FEC) and add a superframe. +---+ |CODEC2 1600 (8 Bytes)

Re: [Freetel-codec2] Codec+ Crypto

I forgot to add the note, that the diagram was for a cipher in counter mode, and that the sequence number and rollover counters were there for that. But a good wiki for encrypting stuff is: https://www.cryptopp.com/wiki/Authenticated_Encryption Which, if you gloss over it, say that encryption is

[Freetel-codec2] coherent modem

rget about whether it is algebra or arithmetic. https://github.com/k5okc/cohfdm75 Have fun, Steve -- ___ Freetel-codec2 mailing list Freetel-codec2@lists.sourceforge

Re: [Freetel-codec2] coherent modem

I was wondering if there was any difference between my modulator and the freedv_tx one, so I created two RAW files and read them into audacity. My first thought was just to mix the two together and play it back with FreeDV. Woof, it actually sounded fine. So the next thought was, to invert one of

Re: [Freetel-codec2] Dumb SM1000 SWD question, where's pin 1?

The one on the bottom goes to R19 and top goes to R20 looks like. I'm looking at the PNG in the repository. Bottom also shows a wider bar on the silkscreen. Bottom being with silkscreen "power" to the left (nearest PTT). --

Re: [Freetel-codec2] Dumb SM1000 SWD question, where's pin 1?

looking at that photo, it's the one on the left. R19 trace sneaks between the right and middle pins and goes down to the left pin. On Sat, Dec 12, 2015 at 12:29 AM, Stuart Longland < stua...@longlandclan.id.au> wrote: > On 12/12/15 16:06, Steve wrote: > > The one on the bo

[Freetel-codec2] encoding travails

I've decided that the hts1.raw file is indestructible and always sounds good :-) I've been trying to find the bug in my java codec2 version, but even with outputting completely different data in the .c2 file, the hts1 voice still sounds good! Course in real-time, my voice sounds like crap, but I'

Re: [Freetel-codec2] encoding travails

dn't be able to get a pitch value of 31 on one machine, and 18 on another :-) When I find it, I'm going to kick myself, I know... On Tue, Dec 15, 2015 at 4:20 PM, Ricardo Andere de Mello wrote: > use bigdecimal? > > 2015-12-15 18:51 GMT-02:00 Steve : >> >> I've d

Re: [Freetel-codec2] encoding travails

Well I got a six-pack of light beer at lunch and worked my way through the analysis part of the codec (pitch and harmonics). I'm OK at that part I think, except for 3 errors of .001 rad/s and 1 of .002. The number of harmonics was the same. So, I guess my quantizer is crapola... I shifted one of

Re: [Freetel-codec2] About the encoded voice quality of SM1000

I'm not sure if any of this is valuable, but on a 64-bit CPU (Pentium G645T) the coherent+700 codec was about: 125k of RAM, which should fit with room to spare for 192 k chips. 75k of Flash or so. (bss needs to be in RAM). I was using 4-byte single-precision floats. textdatabss

Re: [Freetel-codec2] About the encoded voice quality of SM1000

Hmm, left off that the top numbers were for cohpsk, bottom for 700 codec. Also meant to say "NOT much different" on codec sizes. argh... -- Site24x7 APM Insight: Get Deep Visibility into Application Performance APM + Mob

  1   2   3   4   5   6   >