On Sun, Jun 14, 2020 at 12:47:30AM +0200, Michael Niedermayer wrote:
> On Sat, Jun 13, 2020 at 05:11:05PM +0100, Kieran Kunhya wrote:
[...]
> Problem 2
> We do not meassure speed this way because its not realiable nor practical
> just look at this, especially the difference and variation
> ./ffmpeg
>
> The one-tailed (we are not considering whether "fast" is slower than
> "normal") homoscedastic ("equal variance sampled, not paired") student t
> test has a p value of 0.009638597, which for a typical p value threshold of
> 0.05 would reject the null hypothesis ("they are the same speed") and
>
Hi,
On Sat, Jun 13, 2020 at 12:11 PM Kieran Kunhya wrote:
> Hello,
>
> I have run the following commands on the following CPU which is 7 years
> old:
> Intel(R) Xeon(R) CPU E3-1265L v3
>
> ffmpeg_g -i matrixbench_mpeg2.mpg -f null -
> and
> ffmpeg_g -flags2 fast -i matrixbench_mpeg2.mpg -f null
>
> > Also the speed benefits are tiny overall, sub-1%.
>
> This claim is not supported by the statistics you posted
>
For the overall decoding of MPEG-2 the VLC/dequant functions take up
roughly 30% of decoding time.
FAST vs normal is about 2% faster on the VLC/dequant functions, therefore
the to
On Sun, Jun 14, 2020 at 06:50:04PM +0100, Kieran Kunhya wrote:
> On Sun, 14 Jun 2020 at 18:47, Carl Eugen Hoyos wrote:
>
> > Am So., 14. Juni 2020 um 02:33 Uhr schrieb Kieran Kunhya :
> >
> > > Logical fallacy.
> >
> > Could you please stop this?
> >
> > Apart from the personal attack, we have fa
On Sun, Jun 14, 2020 at 06:32:31PM +0100, Kieran Kunhya wrote:
> On Sun, 14 Jun 2020 at 17:32, Michael Niedermayer
> wrote:
>
> > On Sun, Jun 14, 2020 at 05:05:35PM +0100, Kieran Kunhya wrote:
> > > In addition here are results when precomputing the quant tables:
> > >
> > > Precalculated
> > > 6
On Sun, 14 Jun 2020 at 18:47, Carl Eugen Hoyos wrote:
> Am So., 14. Juni 2020 um 02:33 Uhr schrieb Kieran Kunhya :
>
> > Logical fallacy.
>
> Could you please stop this?
>
> Apart from the personal attack, we have far too many known, reproducible
> bugs to spend time on discussing code that you d
Am So., 14. Juni 2020 um 02:33 Uhr schrieb Kieran Kunhya :
> Logical fallacy.
Could you please stop this?
Apart from the personal attack, we have far too many known, reproducible
bugs to spend time on discussing code that you don't like (for whatever
reason) but fortunately don't have to maintai
On Sun, 14 Jun 2020 at 17:32, Michael Niedermayer
wrote:
> On Sun, Jun 14, 2020 at 05:05:35PM +0100, Kieran Kunhya wrote:
> > In addition here are results when precomputing the quant tables:
> >
> > Precalculated
> > 6439
> > 6424
> > 6407
> > 6436
> > 6437
> > Mean 6428.6
> > Stdev 13.42758355
>
On Sun, Jun 14, 2020 at 05:05:35PM +0100, Kieran Kunhya wrote:
> In addition here are results when precomputing the quant tables:
>
> Precalculated
> 6439
> 6424
> 6407
> 6436
> 6437
> Mean 6428.6
> Stdev 13.42758355
does that work on a wider range of videos and CPUs ?
from what i remember really
In addition here are results when precomputing the quant tables:
Precalculated
6439
6424
6407
6436
6437
Mean 6428.6
Stdev 13.42758355
Kieran
On Sun, 14 Jun 2020 at 16:05, Kieran Kunhya wrote:
> Did you remove the branch for FLAGS2_FAST and the large amount of inlined
>> code when making this m
>
> Did you remove the branch for FLAGS2_FAST and the large amount of inlined
> code when making this measurement?
>
I used the attached patch to benchmark this:
Normal Removed
6616 6553
6640 6514
6684 6498
6588 6505
6613 6493
Mean 6628.2 6512.6
Stdev 36.22430124 23.92279248
This means there i
Hi,
> Problem 1
> you have a mean of around 2100 and Stdev of about between 55 and 80 so if
> by
> statistically significant you man 2 Stdev, then with the mean you have.
> You would declare every optimization of less than 6% to be statistically
> insignificant.
> So by what you say here, it seem
On Sat, Jun 13, 2020 at 05:11:05PM +0100, Kieran Kunhya wrote:
> Hello,
>
> I have run the following commands on the following CPU which is 7 years old:
> Intel(R) Xeon(R) CPU E3-1265L v3
>
> ffmpeg_g -i matrixbench_mpeg2.mpg -f null -
> and
> ffmpeg_g -flags2 fast -i matrixbench_mpeg2.mpg -f nul
On 13/06/2020 17:11, Kieran Kunhya wrote:
> It can therefore be shown that the difference is statistically
> insignificant.
I really think the onus of proving FAST is useful should be on
the people who claim it is useful - something I haven't seen
any evidence for.
- Derek
___
15 matches
Mail list logo
