Transcode better than encode - why?

travisbicks · 6th July 2006, 14:45

Until now I have been using DVD Shrink to back up my DVDs so I wanted to try an mpeg-2 encode to see if there would be a noticeable difference in quality. The DVD I chose was the Pal DVD “Fight Club”, the main movie of which is over 7gig. DVD Shrink compressed this at its rating of 58.3%. Using DVD Rebuilder I then encoded the same DVD with QuEnc (0.70) set at two passes at high quality (obtaining a file size of 4gig) and compared the quality between backups. As the only screen I have available at the moment is a 14” CRT monitor, I ran PowerDVD at 4x digital zoom when comparing quality.

I was surprised to find that I preferred the quality of the DVD Shrink backup, as the only discernable difference I could make between them was some “blockiness” in the encoded backup. Could the transcode really be superior in quality to the encode or is my comparison or method flawed?

Daodan · 6th July 2006, 14:52

You said yourself there are blocks in the transcode...do you call that being better? And, no, it can't be better. The only way to make it look better would be a reencode after some filtering in avisynth (and of course, here is largely a matter of taste).

travisbicks · 6th July 2006, 15:00

Sorry, I said that there was blockiness in the encode; I didn't see any in the transcode.

Awatef · 6th July 2006, 15:23

The good thing about DVD Shrink is that it transcodes preserving the highs and lows of the bitrate curve, which is good if the original encode has such highs and lows, which seems to be the case here (well encoded DVD).

But in many cases, the original DVDs are not well encoded, for example with 1-pass encoding only (be it CBR or VBR). In such a case, transcoding à la DVD Shrink would be no good. Here, it is better to use a good MPEG-2 encoder like CCE with 2-pass VBR, so that we get a bitrate curve with "highs and lows", not a relatively flat one, to preserve as much quality as possible.

Daodan · 6th July 2006, 15:26

Now that's a bit weird...at such a difference in size mpeg2 recodes usually display mild to really bad block in high motion areas. Probably the backup was made using the soft option in shrink so probably besides some blocks that disappear, an important amount of detail is also lost.

travisbicks · 6th July 2006, 15:46

Thanks,
I used the default sharp option in shrink. The blocks I've noticed with the encode are in rela
tively still areas of similar colour - close ups of faces for example.

I guess the only way to know which will be better is to do both then

Awatef · 6th July 2006, 17:39

Well, let me tell you this, I tend to use DVDShrink more often because I just don't have the time to play around with CCE & co.
DVDShrink has that 2nd pass option if one hasn't a lot of time but wants to retain as much quality as possible.

But if you do have the time, and wanna see what is the better method, load the VOB in Bitrate Viewer and take a good look at the characteristics of the stream and the bitrate curve. If you see VBR and a beautiful raging curve, you're safe using DVD Shrink, else prepare yourself for a long encoding/authoring procedure

I recommend CCE by the way and nothing else when you have to re-encode. This is the Ferrari under all MPEG-2 encoders... under all aspects, be it speed or quality

travisbicks · 9th July 2006, 15:55

Thanks for the information. I may invest in CCE at some later point when I can afford it. I recently did another DVD Shrink/QuEnc (2pass, high quality, no trellis) comparison with my PAL DVD “The Fifth Element” and found the same thing. A purple wall in the background of one of the shots shows obvious blockiness in the QuEnc version. As my DVDs that I want to backup are feature films, I’m guessing that they’re probably going to be well encoded, so I think I’ll stick to DVD shrink for now.

travisbicks · 11th July 2006, 13:49

Finally got CCE and tried a 3 pass encode of the same DVD. The blocking is not apparent in this encode

Thanks!

ammck55 · 11th July 2006, 23:45

If you read a lot of threads on CCE, you'll probably come to the conclusion that others have come to: more than 2 passes is generally considered to be overkill. Try a 2-pass of one of your recent conversions and compare to your 3-passes and see if the extra time is worth the effort.

ammck55

travisbicks · 15th July 2006, 13:45

Yes, I believe you’re right. I simply wanted to give CCE every advantage when comparing its encode to the results from DVD shrink.

I was quite surprised at the extent of the quality difference between encoding with QuEnc and CCE.

Fishman0919 · 16th July 2006, 05:09

Read this post Link ... it was a test I did awhile back... Shrink is the same but DVD Rebuilder and CCE have got much better

travisbicks · 19th July 2006, 02:32

Thanks, really interesting!

Did you ever get round to comparing CCE with QuEnc? I'd be very interested in the results

Donovan · 20th July 2006, 02:48

Quote:

Originally Posted by Fishman0919

Read this post Link ... it was a test I did awhile back... Shrink is the same but DVD Rebuilder and CCE have got much better

That was a great read, and excellent blind results to your test. Validates what I've found for myself (DVD-RB > *)

Dvd@Best · 20th July 2006, 22:37

DVD shrink gives u 4.7 gb buranble dvdr in just 20mins of time , where are any encoder like CCE or QuEnc would take at least 3 hours to encode a 2 hours movies.Correct me if i am wrong.

I have seen dvds with 100 songs , and was under 4.7gb i wonder how that was encoded.

Dvds with two movies , was also under 4.7gb and quality was not that perfect as it shud be , but it was watchable , i wonder which encoder gives such compressibility

Regards
Dvd@Best

Awatef · 20th July 2006, 23:18

DVDbest, you can always play on the resolution to reduce the bitrate needs to be able to put more video on a disc without sacrifying compression quality. Like if 2 hours are possible with standard 720x480 resolution & 4.8mbps, you could get 4 hours with 352x480 & 2.3mbps, and even 8 hours with 352x240 & 1mbps.
CCE would be the software of choice of course with a decent 2-pass VBR encode (always in MPEG-2 even for 352x240)
And guess what? all these resolutions are DVD-compliant so...

Dvd@Best · 21st July 2006, 00:51

Thats really sound intresting , lets say i wanna backup my dvd with DVD-RB , as its the only one click solution which uses CCE , at wat step i shud stop and change the resolution?

By the way thanks for your positive guideline.

I would also like to know wat resolution is 100% compitable with DVD complaint ,

Regards
Dvd@Best

6th July 2006, 14:45	#1 \| Link
travisbicks Registered User Join Date: Jun 2006 Posts: 22	Transcode better than encode - why? Until now I have been using DVD Shrink to back up my DVDs so I wanted to try an mpeg-2 encode to see if there would be a noticeable difference in quality. The DVD I chose was the Pal DVD “Fight Club”, the main movie of which is over 7gig. DVD Shrink compressed this at its rating of 58.3%. Using DVD Rebuilder I then encoded the same DVD with QuEnc (0.70) set at two passes at high quality (obtaining a file size of 4gig) and compared the quality between backups. As the only screen I have available at the moment is a 14” CRT monitor, I ran PowerDVD at 4x digital zoom when comparing quality. I was surprised to find that I preferred the quality of the DVD Shrink backup, as the only discernable difference I could make between them was some “blockiness” in the encoded backup. Could the transcode really be superior in quality to the encode or is my comparison or method flawed? Last edited by travisbicks; 19th July 2006 at 02:41.

6th July 2006, 14:52	#2 \| Link
Daodan unrecognized user Join Date: Oct 2005 Location: home of Stella Artois Posts: 292	You said yourself there are blocks in the transcode...do you call that being better? And, no, it can't be better. The only way to make it look better would be a reencode after some filtering in avisynth (and of course, here is largely a matter of taste). __________________ Someone once asked me what do I think it's worse, drinking or smoking. I took another drag of my cigarette and I told him I need a drink. GTO

6th July 2006, 15:23	#4 \| Link
Awatef A CORE Media Player User Join Date: May 2002 Location: Africa Posts: 789	The good thing about DVD Shrink is that it transcodes preserving the highs and lows of the bitrate curve, which is good if the original encode has such highs and lows, which seems to be the case here (well encoded DVD). But in many cases, the original DVDs are not well encoded, for example with 1-pass encoding only (be it CBR or VBR). In such a case, transcoding à la DVD Shrink would be no good. Here, it is better to use a good MPEG-2 encoder like CCE with 2-pass VBR, so that we get a bitrate curve with "highs and lows", not a relatively flat one, to preserve as much quality as possible. __________________ Stereo better than 128kbps MP3? Ogg Vorbis at -q2! (~96kbps) / Get the LANCER releases for max conversion speed (at least twice as fast as Lame 3.97!)

6th July 2006, 15:26	#5 \| Link
Daodan unrecognized user Join Date: Oct 2005 Location: home of Stella Artois Posts: 292	Now that's a bit weird...at such a difference in size mpeg2 recodes usually display mild to really bad block in high motion areas. Probably the backup was made using the soft option in shrink so probably besides some blocks that disappear, an important amount of detail is also lost. __________________ Someone once asked me what do I think it's worse, drinking or smoking. I took another drag of my cigarette and I told him I need a drink. GTO

6th July 2006, 15:46	#6 \| Link
travisbicks Registered User Join Date: Jun 2006 Posts: 22	Thanks, I used the default sharp option in shrink. The blocks I've noticed with the encode are in rela tively still areas of similar colour - close ups of faces for example. I guess the only way to know which will be better is to do both then Last edited by travisbicks; 19th July 2006 at 15:00.

6th July 2006, 15:00	#3 \| Link
travisbicks Registered User Join Date: Jun 2006 Posts: 22	Sorry, I said that there was blockiness in the encode; I didn't see any in the transcode.

6th July 2006, 17:39	#7 \| Link
Awatef A CORE Media Player User Join Date: May 2002 Location: Africa Posts: 789	Well, let me tell you this, I tend to use DVDShrink more often because I just don't have the time to play around with CCE & co. DVDShrink has that 2nd pass option if one hasn't a lot of time but wants to retain as much quality as possible. But if you do have the time, and wanna see what is the better method, load the VOB in Bitrate Viewer and take a good look at the characteristics of the stream and the bitrate curve. If you see VBR and a beautiful raging curve, you're safe using DVD Shrink, else prepare yourself for a long encoding/authoring procedure I recommend CCE by the way and nothing else when you have to re-encode. This is the Ferrari under all MPEG-2 encoders... under all aspects, be it speed or quality __________________ Stereo better than 128kbps MP3? Ogg Vorbis at -q2! (~96kbps) / Get the LANCER releases for max conversion speed (at least twice as fast as Lame 3.97!)

9th July 2006, 15:55	#8 \| Link
travisbicks Registered User Join Date: Jun 2006 Posts: 22	Thanks for the information. I may invest in CCE at some later point when I can afford it. I recently did another DVD Shrink/QuEnc (2pass, high quality, no trellis) comparison with my PAL DVD “The Fifth Element” and found the same thing. A purple wall in the background of one of the shots shows obvious blockiness in the QuEnc version. As my DVDs that I want to backup are feature films, I’m guessing that they’re probably going to be well encoded, so I think I’ll stick to DVD shrink for now.

11th July 2006, 13:49	#9 \| Link
travisbicks Registered User Join Date: Jun 2006 Posts: 22	Finally got CCE and tried a 3 pass encode of the same DVD. The blocking is not apparent in this encode Thanks!

11th July 2006, 23:45	#10 \| Link
ammck55 Moderator Join Date: Oct 2002 Location: U.S. of A. Posts: 1,307	If you read a lot of threads on CCE, you'll probably come to the conclusion that others have come to: more than 2 passes is generally considered to be overkill. Try a 2-pass of one of your recent conversions and compare to your 3-passes and see if the extra time is worth the effort. ammck55

15th July 2006, 13:45	#11 \| Link
travisbicks Registered User Join Date: Jun 2006 Posts: 22	Yes, I believe you’re right. I simply wanted to give CCE every advantage when comparing its encode to the results from DVD shrink. I was quite surprised at the extent of the quality difference between encoding with QuEnc and CCE.

16th July 2006, 05:09	#12 \| Link
Fishman0919 Stinkin' up the place!! Join Date: Apr 2004 Location: Somewhere else, maybe Arizona Bay Posts: 1,063	Read this post Link ... it was a test I did awhile back... Shrink is the same but DVD Rebuilder and CCE have got much better __________________ "Cinderella story. Outta nowhere. A former greenskeeper, now, about to become the Masters champion. It looks like a mirac… It's in the hole! It's in the hole! It's in the hole!" Last edited by Fishman0919; 16th July 2006 at 05:14.

19th July 2006, 02:32	#13 \| Link
travisbicks Registered User Join Date: Jun 2006 Posts: 22	Thanks, really interesting! Did you ever get round to comparing CCE with QuEnc? I'd be very interested in the results

20th July 2006, 22:37	#15 \| Link
Dvd@Best Registered User Join Date: May 2002 Location: Asia Posts: 82	the main difference DVD shrink gives u 4.7 gb buranble dvdr in just 20mins of time , where are any encoder like CCE or QuEnc would take at least 3 hours to encode a 2 hours movies.Correct me if i am wrong. I have seen dvds with 100 songs , and was under 4.7gb i wonder how that was encoded. Dvds with two movies , was also under 4.7gb and quality was not that perfect as it shud be , but it was watchable , i wonder which encoder gives such compressibility Regards Dvd@Best

20th July 2006, 23:18	#16 \| Link
Awatef A CORE Media Player User Join Date: May 2002 Location: Africa Posts: 789	DVDbest, you can always play on the resolution to reduce the bitrate needs to be able to put more video on a disc without sacrifying compression quality. Like if 2 hours are possible with standard 720x480 resolution & 4.8mbps, you could get 4 hours with 352x480 & 2.3mbps, and even 8 hours with 352x240 & 1mbps. CCE would be the software of choice of course with a decent 2-pass VBR encode (always in MPEG-2 even for 352x240) And guess what? all these resolutions are DVD-compliant so... __________________ Stereo better than 128kbps MP3? Ogg Vorbis at -q2! (~96kbps) / Get the LANCER releases for max conversion speed (at least twice as fast as Lame 3.97!) Last edited by Awatef; 20th July 2006 at 23:24.

21st July 2006, 00:51	#17 \| Link
Dvd@Best Registered User Join Date: May 2002 Location: Asia Posts: 82	Thats really sound intresting , lets say i wanna backup my dvd with DVD-RB , as its the only one click solution which uses CCE , at wat step i shud stop and change the resolution? By the way thanks for your positive guideline. I would also like to know wat resolution is 100% compitable with DVD complaint , Regards Dvd@Best

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