On Tue, Sep 1, 2015 at 5:54 AM, Robert Krüger wrote:
>> It's been the plan all along to negotiate the removal of the
>> experimental flag after pushing those changes discussed and heavily
>> tested in ticket #2686.
>>
>> If anything, this thread expresses support for that plan ;)
>>
>> So, I'd sti
I'm also looking into it.
It's simply that it tends to get out of sync with the rest really.
I will remove some duplicated code (namely the twoloop and trellis
functions which are verbatim copies of their counterparts in
aaccoder), while somehow keeping the fact that they used the optimized
quant
00:00:00 2001
From: Claudio Freire
Date: Tue, 15 Sep 2015 03:59:45 -0300
Subject: [PATCH] AAC encoder: refactor to resynchronize MIPS port
This patch refactors the AAC coders to reuse code
between the MIPS port and the regular, portable C code.
There were two main functions that had to use
hand-opt
On Tue, Sep 15, 2015 at 8:11 AM, Michael Niedermayer wrote:
> On Tue, Sep 15, 2015 at 04:24:02AM -0300, Claudio Freire wrote:
>> This patch refactors the AAC coders to reuse code
>> between the MIPS port and the regular, portable C code.
>> There were two main functions th
On Wed, Sep 16, 2015 at 12:30 PM, Nedeljko Babic
wrote:
Patch attached.
I thought it was worth a review.
It does include lots of copypaste.
FTR, I tested MIPS 74Kf and x86_64 with make fate-aac
>>>
>>> full fate passes on qemu mips here as well!
>>
>>If
On Fri, Sep 18, 2015 at 3:10 PM, James Almer wrote:
> On 9/15/2015 4:24 AM, Claudio Freire wrote:
>> This patch refactors the AAC coders to reuse code
>> between the MIPS port and the regular, portable C code.
>> There were two main functions that had to use
>>
On Fri, Sep 25, 2015 at 10:39 PM, James Almer wrote:
>> ffmpeg | branch: master | Claudio Freire | Fri
>> Sep 25 03:56:32 2015 -0300| [9458a62decfcaa1313b1ba69276466de536d0768] |
>> committer: Claudio Freire
>>
>> AAC encoder: tweak PNS usage to be more aggr
: -1593.00
From f9159b5514a80f4e19c8b6da108ac9f87272a871 Mon Sep 17 00:00:00 2001
From: Claudio Freire
Date: Sun, 1 Mar 2015 20:11:12 -0300
Subject: [PATCH] Fix AAC Psy PE reduction calculation when multiple iterations
are required AB testing results: http://pastebin.com/051BXYTy
---
libavcodec/aacpsy.c | 2 +-
1 file chang
On Mon, Mar 2, 2015 at 5:50 AM, Carl Eugen Hoyos wrote:
> Claudio Freire gmail.com> writes:
>
>> testing results: http://pastebin.com/051BXYTy
>
> If the testing results are useful please add them
> to the commit message, please do not use links to
> pastebin in the co
Great, picking next patch then.
On Mon, Mar 2, 2015 at 9:48 AM, Michael Niedermayer wrote:
> On Mon, Mar 02, 2015 at 08:47:21AM -0300, Claudio Freire wrote:
>> On Mon, Mar 2, 2015 at 5:50 AM, Carl Eugen Hoyos wrote:
>> > Claudio Freire gmail.com> writes:
>> &g
On Fri, Nov 7, 2014 at 12:08 PM, Nedeljko Babic
wrote:
> Hello,
>
> I am resending patch set with the implementation of fixed point AAC decoder.
>
> Our implementation of float emulation integrated in softfloat as requested.
>
> Patche set rebased to the newest code since there were some changes t
On Tue, Mar 3, 2015 at 10:58 AM, Nedeljko Babic
wrote:
>>On Fri, Nov 7, 2014 at 12:08 PM, Nedeljko Babic
>> wrote:
>>> Hello,
>>>
>>> I am resending patch set with the implementation of fixed point AAC decoder.
>>>
>>> Our implementation of float emulation integrated in softfloat as requested.
>>>
df43314a67ad625775795561d960c30408fcf892 Mon Sep 17 00:00:00 2001
From: Claudio Freire
Date: Fri, 6 Mar 2015 04:05:32 -0300
Subject: [PATCH] AAC: Add support for 7350Hz sampling rates, no error on too
hight bitrate.
Instead, warn that bitrate will be clamped down to the maximum allowed.
Patch is
On Fri, Mar 6, 2015 at 9:59 AM, Michael Niedermayer wrote:
> On Fri, Mar 06, 2015 at 04:28:38AM -0300, Claudio Freire wrote:
>> AAC: Add support for 7350Hz sampling rates, no error on too hight bitrate.
>>
>> Instead, warn that bitrate will be clamped down to the maximum al
On Fri, Mar 6, 2015 at 11:47 AM, Michael Niedermayer wrote:
>> 2- How to run that, make check dies at dct1d-4
>
> make check should pass, if it does not please try without local
> changes, with make distclean or if (unlikely) these still fail a
> fresh git checkout
I did use a clean checkout.
Ju
not just any samples, it wants
>> a specific set of samples.
>
> configure with --samples= to point to some directory and run
> make fate-rsync
> (this will need about 900mb space)
Yeah, that works.
Attached another patch that adds the tests. Goes on top of the initial
patch (wit
On Sun, Mar 8, 2015 at 8:23 AM, Michael Niedermayer wrote:
> i think its better to leave fuzz at a small value otherwise we
> would forget to update the target when it improves and then
> subsequent worsenings could be missed
An updated version with that fixed.
It's a bit tricky to squash commit
On Sun, Mar 8, 2015 at 11:06 PM, Claudio Freire wrote:
> On Sun, Mar 8, 2015 at 8:23 AM, Michael Niedermayer wrote:
>> i think its better to leave fuzz at a small value otherwise we
>> would forget to update the target when it improves and then
>> subsequent worsenings c
On Mon, Mar 9, 2015 at 12:41 PM, Michael Niedermayer wrote:
> On Sun, Mar 08, 2015 at 11:07:33PM -0300, Claudio Freire wrote:
>> On Sun, Mar 8, 2015 at 11:06 PM, Claudio Freire
>> wrote:
>> > On Sun, Mar 8, 2015 at 8:23 AM, Michael Niedermayer
>> > wrote:
On Mon, Mar 9, 2015 at 1:05 PM, Michael Niedermayer wrote:
>> Head is still using a butterworth filter for the lowpass IIRC. That
>> could very well be both platform and sample rate specific.
>>
>> Sadly, I cannot test mips, but commenting the lowpass could yield
>> useful results.
>
> what exact
On Mon, Mar 9, 2015 at 1:26 PM, Michael Niedermayer wrote:
> doesnt help
> but
> this does:
> -fate-aac-s7350-encode: CMD = enc_dec_pcm adts wav s16le $(REF) -strict -2
> -c:a aac -b:a 256k
> +fate-aac-s7350-encode: CMD = enc_dec_pcm adts wav s16le $(REF) -strict -2
> -c:a aac -b:a 88k
>
> this
On Mon, Mar 9, 2015 at 1:38 PM, Michael Niedermayer wrote:
> On Mon, Mar 09, 2015 at 01:31:26PM -0300, Claudio Freire wrote:
>> On Mon, Mar 9, 2015 at 1:26 PM, Michael Niedermayer wrote:
>> > doesnt help
>> > but
>> > this does:
>> > -fate-aac-s7350-e
On Mon, Mar 9, 2015 at 3:31 PM, Claudio Freire wrote:
> On Mon, Mar 9, 2015 at 1:38 PM, Michael Niedermayer wrote:
>> On Mon, Mar 09, 2015 at 01:31:26PM -0300, Claudio Freire wrote:
>>> On Mon, Mar 9, 2015 at 1:26 PM, Michael Niedermayer
>>> wrote:
>>> >
On Tue, Mar 10, 2015 at 8:05 AM, Michael Niedermayer wrote:
>> I cannot test myself yet (cannot build against mips, it will
>> probably take me a long while before I can set up qemu for that), but
>> the patch should bring both implementations in sync, at least enough
>> for the test to pass witho
On Tue, Mar 10, 2015 at 10:08 AM, Claudio Freire wrote:
> On Tue, Mar 10, 2015 at 8:05 AM, Michael Niedermayer wrote:
>>> I cannot test myself yet (cannot build against mips, it will
>>> probably take me a long while before I can set up qemu for that), but
>>&g
On Fri, Mar 13, 2015 at 2:34 PM, Claudio Freire wrote:
> On Fri, Mar 13, 2015 at 12:39 PM, Nedeljko Babic
> wrote:
>>> btw, i use the qemu from https://github.com/ssvb/QEMU.git
>>> for mips as it at least years ago supported more extensions, i
>>> would have gue
On Fri, Mar 13, 2015 at 12:39 PM, Nedeljko Babic
wrote:
>> btw, i use the qemu from https://github.com/ssvb/QEMU.git
>> for mips as it at least years ago supported more extensions, i
>> would have guessed these where merged into main qemu but as you list
>> all the disables in the wiki, maybe i wa
On Fri, Mar 13, 2015 at 3:06 PM, Michael Niedermayer wrote:
> On Fri, Mar 13, 2015 at 02:34:08PM -0300, Claudio Freire wrote:
>> On Fri, Mar 13, 2015 at 12:39 PM, Nedeljko Babic
>> wrote:
>> >> btw, i use the qemu from https://github.com/ssvb/QEMU.git
>> >
On Mon, Mar 16, 2015 at 12:41 AM, Michael Niedermayer wrote:
> On Fri, Mar 13, 2015 at 03:53:56PM -0300, Claudio Freire wrote:
>> On Fri, Mar 13, 2015 at 3:06 PM, Michael Niedermayer
>> wrote:
>> > On Fri, Mar 13, 2015 at 02:34:08PM -0300, Claudio Freire wrote:
>>
On Sun, Apr 12, 2015 at 1:50 AM, Rostislav Pehlivanov
wrote:
> +int diff, off_sf = sce->sf_idx[0], off_pns = sce->sf_idx[0];
If you take a look at the decoder, off_pns has to be initialized as off_pns - 90
___
ffmpeg-devel mailing list
ffmpeg-devel
On Sun, Apr 12, 2015 at 11:49 PM, Claudio Freire wrote:
> On Sun, Apr 12, 2015 at 1:50 AM, Rostislav Pehlivanov
> wrote:
>> +int diff, off_sf = sce->sf_idx[0], off_pns = sce->sf_idx[0];
>
>
> If you take a look at the decoder, off_pns has to be initialized as o
On Sun, Apr 12, 2015 at 1:50 AM, Rostislav Pehlivanov
wrote:
> @@ -245,6 +258,8 @@ static float (*const quantize_and_encode_band_cost_arr[])(
> quantize_and_encode_band_cost_UPAIR,
> quantize_and_encode_band_cost_UPAIR,
> quantize_and_encode_band_cost_ESC,
> +NULL,
> +quanti
Looks good.
Though it does depend on the first patch of the series.
On Sun, Apr 12, 2015 at 1:50 AM, Rostislav Pehlivanov
wrote:
> This commit changes the decoder to use the macro definitions for scalefactors
> and noise energy values reading. This was done to reduce confusion when
> looking a
On Sun, Apr 12, 2015 at 11:52 PM, Claudio Freire wrote:
> On Sun, Apr 12, 2015 at 11:49 PM, Claudio Freire
> wrote:
>> On Sun, Apr 12, 2015 at 1:50 AM, Rostislav Pehlivanov
>> wrote:
>>> +int diff, off_sf = sce->sf_idx[0], off_pns = sce->sf_idx[0];
&
On Mon, Apr 13, 2015 at 12:02 AM, Claudio Freire wrote:
> Looks good.
>
> Though it does depend on the first patch of the series.
>
>
> On Sun, Apr 12, 2015 at 1:50 AM, Rostislav Pehlivanov
> wrote:
>> This commit changes the decoder to use the macro definitions for
LGTM
On Mon, Apr 13, 2015 at 8:33 PM, Rostislav Pehlivanov
wrote:
> This commit replaces the previous hardcoded constants with both new and
> previously defined macros from aac.h. This change makes it easy for anyone
> reading the code to know how encoding and decoding scalefactors works. It's
t; uses additional codebooks.
>>
>> The 12th entry in the codebook function array points to a function which
>> stops the execution of the program by calling an assert with an always
>> 'false' argument. After a discussion, it was pointed out in an email
>>
ion of the program by calling an assert with an always
> 'false' argument. After a discussion, it was pointed out in an email
> discussion with Claudio Freire that having a 'NULL' entry can result in
> unexpected behaviour and could be used as a security hole. Ther
On Tue, Apr 14, 2015 at 4:52 PM, Rostislav Pehlivanov
wrote:
> Uhh, can't replicate bug here (freshly built ffmpeg, just applied this patch
> only), (md5 for file = 473edd68b91123c3a9c1825271012357). tried other files
> and they encode and play fine. Spectrum also looks fine. I also tested other
>
It's close...
On Tue, Apr 14, 2015 at 11:17 PM, Rostislav Pehlivanov
wrote:
> @@ -711,9 +748,11 @@ static void search_for_quantizers_twoloop(AVCodecContext
> *avctx,
> {
> int start = 0, i, w, w2, g;
> int destbits = avctx->bit_rate * 1024.0 / avctx->sample_rate /
> avctx->channels *
On Wed, Apr 15, 2015 at 8:18 AM, Rostislav Pehlivanov
wrote:
> @@ -828,9 +882,17 @@ static void search_for_quantizers_twoloop(AVCodecContext
> *avctx,
>
> fflag = 0;
> minscaler = av_clip(minscaler, 60, 255 - SCALE_MAX_DIFF);
> +
> +for (w = 0; w < sce->ics.num_windows;
On Wed, Apr 15, 2015 at 1:29 PM, Nedeljko Babic
wrote:
> Hello,
>
> I am again resending patch set with the implementation of fixed point AAC
> decoder.
>
> Patch set is rebased to the newest code as requested.
Cool, I'll test them when I get some free time, hopefully during this week.
On Thu, Apr 16, 2015 at 1:00 PM, Andreas Cadhalpun
wrote:
> If both band->active_lines and band->thr are 0.0f, the division is
> undefined, making norm_fac not a number.
>
> NaN is passed on to other variables until it finally reaches
> sce->sf_idx and is converted to an integer (-2147483648).
>
>
On Thu, Apr 16, 2015 at 3:19 PM, Andreas Cadhalpun
wrote:
> On 16.04.2015 19:41, Claudio Freire wrote:
>> It should be if band->thr > 0.0f, all divisions by zero return
>> something that casts into an ~1:
>
> OK, new patch attached.
>
> If band->thr is 0.0f and
On Sat, Apr 18, 2015 at 10:34 AM, Michael Niedermayer wrote:
> On Sat, Apr 18, 2015 at 01:59:53PM +0200, Andreas Cadhalpun wrote:
>> On 18.04.2015 04:40, Michael Niedermayer wrote:
>> > On Sat, Apr 18, 2015 at 12:55:08AM +0200, Andreas Cadhalpun wrote:
>> >> The problem is that minath is not the m
On Mon, Apr 20, 2015 at 7:18 PM, Andreas Cadhalpun
wrote:
> On 20.04.2015 22:15, Claudio Freire wrote:
>> TL;DR, band->thr should not be negative ever, band->thr == 0.0f would
>> cause lots of issues on its own if band->energy != 0.0f in such a case
>> (though I
On Mon, Apr 20, 2015 at 8:32 PM, Andreas Cadhalpun
wrote:
The long version:
ath should approximate the shape of the absolute hearing threshold, so
yes, it's best if it really uses the minimum, since that will prevent
clipping of the ath curve and result in a more accurate
On Mon, Apr 20, 2015 at 8:59 PM, Claudio Freire wrote:
> On Mon, Apr 20, 2015 at 8:32 PM, Andreas Cadhalpun
> wrote:
>>>>> The long version:
>>>>>
>>>>> ath should approximate the shape of the absolute hearing threshold, so
>>>>
On Mon, Apr 20, 2015 at 9:13 PM, Michael Niedermayer wrote:
> On Mon, Apr 20, 2015 at 09:07:14PM -0300, Claudio Freire wrote:
>> On Mon, Apr 20, 2015 at 8:59 PM, Claudio Freire
>> wrote:
>> > On Mon, Apr 20, 2015 at 8:32 PM, Andreas Cadhalpun
>> >
On Wed, Apr 22, 2015 at 10:23 AM, Andreas Cadhalpun
wrote:
> +if (k == last_k && msb == last_msb) {
> +av_log(ac->avctx, AV_LOG_ERROR, "patch construction failed\n");
> +return AVERROR_INVALIDDATA;
> +}
> +last_k = k;
> +last_msb = msb;
I d
On Wed, Apr 22, 2015 at 12:54 PM, Andreas Cadhalpun
wrote:
> On 22.04.2015 17:35, Claudio Freire wrote:
>> On Wed, Apr 22, 2015 at 10:23 AM, Andreas Cadhalpun
>> wrote:
>>> +if (k == last_k && msb == last_msb) {
>>> +av_log(ac->avc
On Wed, Apr 22, 2015 at 1:52 PM, Vittorio Giovara
wrote:
I don't think the INVALIDDATA return will have the desired effect.
I think you want return -1;
>>>
>>> This function is only called once and there the check is:
>>> if (sbr_hf_calc_npatches(ac, sbr) < 0)
>>> return
On Thu, Apr 23, 2015 at 12:43 PM, Michael Niedermayer wrote:
> On Thu, Apr 23, 2015 at 12:22:49PM -0300, Claudio Freire wrote:
>> On Wed, Apr 22, 2015 at 1:52 PM, Vittorio Giovara
>> wrote:
>> >>>> I don't think the INVALIDDATA return will have the desir
101 - 153 of 153 matches
Mail list logo
