I see in the Makefile that __NO_MATH_INLINES was added to fix a bug in gcc
2.8+, what bug is this, what does this do and why do you have it defined?
And what is RH_AMP?
Josh
----- Original Message -----
From: "Robert Hegemann" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Wednesday, August 16, 2000 1:40 PM
Subject: Re: Re[2]: [MP3 ENCODER] compiling plug-ins w/ Cygwin gcc
Hi Dimitry and Nathan!
Nathan Blomquist schrieb am Mit, 16 Aug 2000:
> I did some testing.
>
> I had a test.wav and I used my lame-3_86BetaWin32.zip and my
lame-3_86BetaDJGPP.zip.
>
> I then tried -v only. They came up different. I then tried --vbr-new and
> they both came out the same. Then came --vbr-old. This is where it got
> interesting. DJGPP gave me an avg of 181 kbs while VC gave me 120 kbs. I
did
> not change any code in any of the files when I built these two, so I have
no
> idea why the drastic change.
>
> Thanks for pointing this out.
I did a comparison between GCC 2.95.2 under Linux
and Intel Compiler 4.5 under Windows.
I compiled todays LAME (fresh from CVS) with
GCC 2.95.2:
- MMX_choose_table on
- RH_AMP on
- MATH_INLINES enabled
- Compiler flags as for gcc-2.95.2 (uncommented them)
Intel Compiler 4.5:
- used Makefile.MSVC (as I checked it in today)
I used "lame -v <filename>":
here are the results for a bunch of test songs under Linux:
52 frames 86 kbit/s 0:01.352 - 45st.wav.mp3
194 frames 33 kbit/s 0:05.044 - 60.wav.mp3
318 frames 180 kbit/s 0:08.268 - applaud.wav.mp3
271 frames 120 kbit/s 0:07.046 - BarbaraAnn.wav.mp3
256 frames 148 kbit/s 0:06.656 - castanets.wav.mp3
331 frames 113 kbit/s 0:08.606 - clips.wav.mp3
219 frames 133 kbit/s 0:05.694 - else3.wav.mp3
194 frames 216 kbit/s 0:05.044 - fatboy.wav.mp3
1187 frames 134 kbit/s 0:30.862 - Fools.wav.mp3
804 frames 144 kbit/s 0:20.904 - Franco.wav.mp3
982 frames 158 kbit/s 0:25.532 - ftb_samp.wav.mp3
270 frames 100 kbit/s 0:07.020 - hippies.wav.mp3
1341 frames 141 kbit/s 0:34.866 - iron.wav.mp3
307 frames 149 kbit/s 0:07.982 - kingdom.wav.mp3
232 frames 136 kbit/s 0:06.032 - KMFDM-Dogma.wav.mp3
378 frames 166 kbit/s 0:09.828 - Main Theme.cut.wav.mp3
83 frames 96 kbit/s 0:02.158 - masking.wav.mp3
154 frames 139 kbit/s 0:04.004 - mstest.wav.mp3
385 frames 99 kbit/s 0:10.010 - ns.wav.mp3
306 frames 108 kbit/s 0:07.956 - t1.wav.mp3
55 frames 33 kbit/s 0:01.430 - teenyweenytinybug.wav.mp3
181 frames 119 kbit/s 0:04.706 - testclip.wav.mp3
157 frames 128 kbit/s 0:04.082 - testsignal2.wav.mp3
174 frames 121 kbit/s 0:04.524 - testsignal4.wav.mp3
148 frames 150 kbit/s 0:03.848 - track7.wav.mp3
457 frames 179 kbit/s 0:11.882 - velvet.wav.mp3
and here under Windows:
52 frames 86 kbit/s 0:01.352 - 45st.wav.mp3
194 frames 33 kbit/s 0:05.044 - 60.wav.mp3
318 frames 180 kbit/s 0:08.268 - applaud.wav.mp3
271 frames 120 kbit/s 0:07.046 - BarbaraAnn.wav.mp3
256 frames 148 kbit/s 0:06.656 - castanets.wav.mp3
331 frames 113 kbit/s 0:08.606 - clips.wav.mp3
219 frames 133 kbit/s 0:05.694 - else3.wav.mp3
194 frames 216 kbit/s 0:05.044 - fatboy.wav.mp3
1187 frames 134 kbit/s 0:30.862 - Fools.wav.mp3
804 frames 144 kbit/s 0:20.904 - Franco.wav.mp3
982 frames 158 kbit/s 0:25.532 - ftb_samp.wav.mp3
270 frames 100 kbit/s 0:07.020 - hippies.wav.mp3
1341 frames 141 kbit/s 0:34.866 - iron.wav.mp3
307 frames 149 kbit/s 0:07.982 - kingdom.wav.mp3
232 frames 136 kbit/s 0:06.032 - KMFDM-Dogma.wav.mp3
378 frames 166 kbit/s 0:09.828 - Main Theme.cut.wav.mp3
83 frames 96 kbit/s 0:02.158 - masking.wav.mp3
154 frames 139 kbit/s 0:04.004 - mstest.wav.mp3
385 frames 99 kbit/s 0:10.010 - ns.wav.mp3
306 frames 108 kbit/s 0:07.956 - t1.wav.mp3
55 frames 33 kbit/s 0:01.430 - teenyweenytinybug.wav.mp3
181 frames 119 kbit/s 0:04.706 - testclip.wav.mp3
157 frames 128 kbit/s 0:04.082 - testsignal2.wav.mp3
174 frames 121 kbit/s 0:04.524 - testsignal4.wav.mp3
148 frames 150 kbit/s 0:03.848 - track7.wav.mp3
457 frames 179 kbit/s 0:11.882 - velvet.wav.mp3
as I could not see any difference here, I compared the file sizes:
robert@bob:~ > diff A.dir B.dir
13c13
< -rw-rw-rw- 1 nobody nogroup 136474 Aug 16 16:35 fatboy.wav.mp3
---
> -rw-rw-rw- 1 nobody nogroup 136420 Aug 16 16:35 fatboy.wav.mp3
20c20
< -rw-rw-rw- 1 nobody nogroup 124455 Aug 16 16:40 ns.wav.mp3
---
> -rw-rw-rw- 1 nobody nogroup 124510 Aug 16 16:40 ns.wav.mp3
As you can see, only fatboy and ns differ a few bytes in file size,
due to some different roundings. All other files are binary identical.
So I don't know why you both get such different results with DJGPP.
Ciao Robert
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
