tmpgenc 2.5 plus does not encode at chosen VBR bitrate

kds2012 · 26th August 2009, 02:40

Has anybody had this problem using tmpgenc 2.5:

You set the 2-pass VBR Average rate at a specific value,
but the output .m2v is not encoded at that rate?
Specifically, it is encoded at a lower rate/quality than
specified, and the file size is less than expected?

This happens intermittently on my XP x64 system --
I haven't detected a pattern yet. Has anybody else seen this??

Blue_MiSfit · 28th August 2009, 00:36

Dunno... TMPGEnc 2.5 is _extremely_ old. It might have bizarre problems like this on newer systems. I'd say try out another MPEG-2 encoder, like HC or QuEnc.

~MiSfit

Taurus · 28th August 2009, 13:04

Maybe the encoder gets saturated.
Out of curiosity:
Lower the bitrate settings avr/max/min to same settings.
For a test: 1500 kbits/sec
And you will see TmpgEnc will hit the desired bitrate.
Without knowing any of your sources/bitrate/resolution it's only guessing.
Or maybe your settings are messed up. Different Quantize Matrix? etc.....
Set everything to default, or delete the Tmpgenc.ini.
Warning, this will zero any of your customations to default.
Just verified my suggestions and Mediainfo shows exact numbers at lower bitrates 1500/2000 kbits/sec.
But at 8000 kbits/sec encoder gets saturated to about 4500 kbits/sec at this source.
This is on XP 32bit....
Hope this helps

kds2012 · 28th August 2009, 23:15

Hey thx Taurus -- saturation seems to be the issue -- and it's not one I'm familiar with. Care to share what you know? (I'll be searching elsewhere in any case.)

kds2012 · 29th August 2009, 01:58

I think I get it now... on VBR at least, the encoder won't exceed the source -- I always figured it was like MP3 -- if you told it to oversample it would, even if there was no quality gain.

26th August 2009, 02:40	#1 \| Link
kds2012 Registered User Join Date: Jun 2008 Posts: 5	tmpgenc 2.5 plus does not encode at chosen VBR bitrate Has anybody had this problem using tmpgenc 2.5: You set the 2-pass VBR Average rate at a specific value, but the output .m2v is not encoded at that rate? Specifically, it is encoded at a lower rate/quality than specified, and the file size is less than expected? This happens intermittently on my XP x64 system -- I haven't detected a pattern yet. Has anybody else seen this??

28th August 2009, 00:36	#2 \| Link
Blue_MiSfit Derek Prestegard IRL Join Date: Nov 2003 Location: Los Angeles Posts: 5,993	Dunno... TMPGEnc 2.5 is _extremely_ old. It might have bizarre problems like this on newer systems. I'd say try out another MPEG-2 encoder, like HC or QuEnc. ~MiSfit __________________ These are all my personal statements, not those of my employer :)

29th August 2009, 01:58	#5 \| Link
kds2012 Registered User Join Date: Jun 2008 Posts: 5	I see .. I think I get it now... on VBR at least, the encoder won't exceed the source -- I always figured it was like MP3 -- if you told it to oversample it would, even if there was no quality gain.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

28th August 2009, 13:04	#3 \| Link
Taurus Registered User Join Date: Mar 2002 Location: Krautland Posts: 903	Maybe the encoder gets saturated. Out of curiosity: Lower the bitrate settings avr/max/min to same settings. For a test: 1500 kbits/sec And you will see TmpgEnc will hit the desired bitrate. Without knowing any of your sources/bitrate/resolution it's only guessing. Or maybe your settings are messed up. Different Quantize Matrix? etc..... Set everything to default, or delete the Tmpgenc.ini. Warning, this will zero any of your customations to default. Just verified my suggestions and Mediainfo shows exact numbers at lower bitrates 1500/2000 kbits/sec. But at 8000 kbits/sec encoder gets saturated to about 4500 kbits/sec at this source. This is on XP 32bit.... Hope this helps

28th August 2009, 23:15	#4 \| Link
kds2012 Registered User Join Date: Jun 2008 Posts: 5	Hey thx Taurus -- saturation seems to be the issue -- and it's not one I'm familiar with. Care to share what you know? (I'll be searching elsewhere in any case.)