|
Does DVD-RB 'process' files even at 0% compression like DVDShrink does...
|
|
Staff Member
2 product reviews
|
6. October 2005 @ 18:44 |
Link to this message
|
|
Pixels are always in the same location. What SSIM measures is the actual luma and chroma values of those pixels. So basically, no they couldn't be any different if it says they're the same. Based on my own findings I'd be very interested to have anyone show me a single frame that Shrink says it didn't compress but is different in any way from the original.
|
|
Advertisement
|
|
|
|
|
Lumbastio
Junior Member
|
6. October 2005 @ 22:02 |
Link to this message
|
|
I just noticed that you had finally released your test results. Thanks for taking this matter to the end vurbal. As I said, I'm not a pro in DVD copying, but what you say makes sense. I can only back that up with a qualitative analysis - I used Shrink to edit a DVD in an uncompressed format and then processed though CCE SP 4 passes. I couldn't notice any difference between the original and the copy (Sony Trinitron 61 or so inches.) I think that if someone disagrees with the your conlusions, they should publish their own test results that clearly show that Shrink reduces the quality, otherwise discussions don't help much without any empirical data. If you go deep enough, you can successfully argue that the copy is not identical to the original. Bu then again, you can use the same argument to argue that no two original DVD's are exactly the same since their surfaces don't all have the exact same 3D contours burned in by the photons in the laser of the burners in the factory. Does this mean that this is relevant? No. It can't be perceptible by humans and can be left out. Believe me, in physics they make much rougher approximation when they build a spaceship or the equipment for your DVD burner. If we go any deeper, we'll have to decide whether God decides to hide the extra kb to return the energy to the universe as entropy. Forget it.
XP SP2
512MB RAM
Pentium IV CPU 1.80GHz
NEC 3540A
BEST QUALITY DVD BURNING (in order of function):
1) AnyDVD 5.4
2) DVD Decrypter
3) DVD Rebuilder 1.00RC60Pro
4) CCE SP
5) Nero 6.6 Ultra
[InterVideo DVD Copy 4]
This message has been edited since posting. Last time this message was edited on 6. October 2005 @ 22:04
|
AfterDawn Addict
1 product review
|
8. October 2005 @ 08:27 |
Link to this message
|
|
I've done some studying of SSIM and it doesn't do exact comparisons it only does similarities. But that being said I agree with Lumbastio in that the majoriy of claims made by others regading Shrink are qualitative measures. Therefore I've down loaded SSIM and I'm going to step back and for a while and do my own tests.
" Please Read!!! Post your questions only in This Thread or they will go unanswered:
Help with development of BD RB: Donations at: http://www.jdobbs.com/. This message has been edited since posting. Last time this message was edited on 14. October 2005 @ 13:04
|
Staff Member
2 product reviews
|
8. October 2005 @ 09:10 |
Link to this message
|
One thing I hadn't considered which may be of importance here is the weighting of different planes. Since the weight of the chroma planes is only .1 for each compared to .8 for the lume it's possible that luma differences could be small enough to disappear due to lack of significant digits in the calculations. That means the only truly reliable measurement would be a straight binary comparison with a hex editor to verify they're identical. The downside to this method is that it may show differences in the header that don't make any difference in the actual visual information, but as long as I Demux both versions with the same software that shouldn't be a problem.
|
AfterDawn Addict
1 product review
|
8. October 2005 @ 09:16 |
Link to this message
|
vurbal
Give me a break. LOL
I started writing AVI script a year ago and then I got sidetracked doing other things. Now I'm back to studying it again so I might be sending some question your way.
From what I've read the value 1 denotes the similarity.
I may prove myself wrong but I still believe that DVD Shrink somehow filters even a ripped no compression result. From what I've read SSIM does a pretty good job of making comparisons.
Question does it also compare bitrates?
" Please Read!!! Post your questions only in This Thread or they will go unanswered:
Help with development of BD RB: Donations at: http://www.jdobbs.com/. |
Staff Member
2 product reviews
|
8. October 2005 @ 09:21 |
Link to this message
|
|
It doesn't compare bitrates directly but, assuming a value of 1 actually shows that the planes are identical, the bitrate would also have to be identical. Of course a comparison using Bitrate Viewer or MPEG Stream Eye would be another test that could verify that.
Edit: No breaks here ;) There are always more questions to be asked
This message has been edited since posting. Last time this message was edited on 8. October 2005 @ 09:23
|
AfterDawn Addict
1 product review
|
8. October 2005 @ 09:58 |
Link to this message
|
Vurbal
Quote:
There are always more questions to be asked
For that, we can all be thankful. Without questions there would be nothing to learn.
I have a 4 day weekend coming up, I might be able to give it go.
" Please Read!!! Post your questions only in This Thread or they will go unanswered:
Help with development of BD RB: Donations at: http://www.jdobbs.com/. |
Staff Member
2 product reviews
|
8. October 2005 @ 10:14 |
Link to this message
|
|
Let me know if you need any help with SSIM.
|
AfterDawn Addict
1 product review
|
8. October 2005 @ 10:18 |
Link to this message
|
|
vurbal
Thanks! I will.
" Please Read!!! Post your questions only in This Thread or they will go unanswered:
Help with development of BD RB: Donations at: http://www.jdobbs.com/. |
|
Lumbastio
Junior Member
|
8. October 2005 @ 20:12 |
Link to this message
|
|
Good luck Sophocles. Let's hope that our suspicions will finally be put to rest by confirming in this thread that the uncopressed format of Shrink does not alter the edited DVD quality. Remember, that if you find infinitesimally small differences between the original DVD and the one processed by Shrink, there will still be the issue whether that small difference in numbers can also be distinguished visually. So if your meter stick is DVD Remake, in that case you might have to analyze the same DVD processed by Remake to be able to compare it to Shrink becase in case of a very small difference it might turn out that all programs affect the DVD in the same manner. Here is a suggestion: for very subtle differences, small errors are going to have huge effects on the results. So it would be best if you compare the processed copy not to the DVD directly, but to the exact same file ripped onto your hard drive. This way you will eliminate the small differences that your burning hardware creates evertime it copies a DVD. Again, if the differences are going to be large, this won't matter, but it won't hurt.
XP SP2
512MB RAM
Pentium IV CPU 1.80GHz
NEC 3540A
BEST QUALITY DVD BURNING (in order of function):
1) AnyDVD 5.4
2) DVD Decrypter
3) DVD Rebuilder 1.00RC60Pro
4) CCE SP
5) Nero 6.6 Ultra
[InterVideo DVD Copy 4]
|
AfterDawn Addict
1 product review
|
8. October 2005 @ 20:47 |
Link to this message
|
Lumbastio
It's already known that DVD Shrink under no compression renders a small difference in physical size. You can verify that by ripping a DVD with Decrypter and then running that same ripped file through Shrink targeted to another folder and then compare the two. I've also had compressed DVDs that weren't compliant but could be made so just by running it through DVD Shrink. So there are already small differences between the original and the no compressed output but the question is does it affect picture quallity in some way.
" Please Read!!! Post your questions only in This Thread or they will go unanswered:
Help with development of BD RB: Donations at: http://www.jdobbs.com/. |
|
Advertisement
|
|
|
|
brewha
Suspended due to non-functional email address
|
3. February 2006 @ 10:10 |
Link to this message
|
|
Sorry for bumping this super old thread, but has there been a final conclusion to the debate about Shrink altering files.
|
