MichaelG
Second Unit
- Joined
- Jul 10, 2000
- Messages
- 322
Gary, that is intersting. I guess that would answer my question as well.
Well, over-compressing an image with a DCT technique like MPEG2 or JPEG will reduce the effects of edge enhancement, since one of the stages in M(J)PEG compression is a low-pass filter.
I'm skeptical of this explanation, unless you can show me that low bit-rate DVDs (more compression) have less EE than high bit rate DVDs (less compression).
As I said, the theatrical digital presentation I saw (and which obviously had much LESS compression than ANY DVD) had ZERO EE. Which tells me that the "source" argument doesn't hold.
The other thing I'm having a problem with this issue in general is that everyone continues to use the term "EE", meaning "edge enhancement". The visible artifacts is actually "ringing", which can be caused by digital over-enhancement.
Robert, I think the semantic use is irrelevant. Whether you choose to call it "digitally oversharpened" or "edge enhanced", it amounts to the same thing, ie, a process that studios should not be using. Here's a quote from Bjoern's guide to EE:
"If you have ever increased the sharpness control on your TV to the max, you will also get the same ringing, so you are well advised to keep it as low as possible. The advantage of controlling the subjective sharpness as a control on the TV, instead of having the DVD transfer sharpened through the use of EE, is that you can adjust the sharpness control on your TV to your liking. But you can't get the video-like look out of a DVD transfer that has EE applied to it! You have to live with it, its an inreversible process.
Note: the ringing that is caused by setting the sharpness control too high on your TV is only a 1-dimensional horizontal effect. It show ringing only at left and right of vertical edges. EE on the other hand is applied in two dimesions, so both horizonal and vertical edges can show ringing. But EE is not necessary a symmetrical filter as we will see. The amount of horizontal and vertical EE used on DVDs can vary considerably.
Does 'Edge Enhancement' always look the same?
No, there are several parameters that affect the outcome of an EE processed image. First of all, the amount of EE applied can vary and will result in differnt amplitudes of the modulation. Little EE will result in only faintly visible halos, lots of EE will result in stronly visible, high contrast contours. Secondly, the frequency of the filter can vary. A high frequency filter will render only thin halos, while a low frequency one will render thick halos. Thirdly, the symmetry of the filter can vary. Stronger horizontally than vertically, for example."
Watching the film on both my 27" television and on a computer display (which can take advantage of the anamorphic enhancement), I noticed zero problems with the transfer. Despite the extreme amount of digital work on the film and that it was presented in digital theaters, the DVD transfer was transferred from film. It has an amazing filmlike quality; very rarely can you spot scenes that look digital, which is an important accomplishment for both the Industrial Light and Magic effects artists and the technicians who performed the transfer.
I was going to e-mail the reviewer about it, but on second thought why should I spoil it for him. He's happy.
Columbia / Tristar is very consistent. They always use it on 2.35:1 transfers (recently 6th Day) and don't on 1.85:1 transfer (Starship Troopers is still one of the best non-EE transfers out there).
Heres my thinking: If only the 2.35 transfers have the problem, then I do not think the Sony encoder is to blame- assuming CT uses the same encoder for all of thier titles (which I think is pretty safe to assume) Could it possibly be related to the electroniv Anamorphic process of unsqueezing the 2.35 film element and them resqeezing it for the 16x9 anamorphic video format? Maybe that would explain the fact that it is assymetrical? It might also explain why Titanic (in super35) has none while Star Wars Ep I (in arriscope) does?
Am I way off target?
-cris
If only the 2.35 transfers have the problem, then I do not think the Sony encoder is to blame- assuming CT uses the same encoder for all of thier titles (which I think is pretty safe to assume) Could it possibly be related to the electroniv Anamorphic process of unsqueezing the 2.35 film element and them resqeezing it for the 16x9 anamorphic video format?
Am I way off target?
You are off target Cris, because you're assuming that CT doesn't use the 16:9 anamorphic process on their 1.85 transfers.
In fact, they DO use it.
To me, this just proves how UNNECESSARY EE is.
I don't know if it's on TPM
Again, I don't know if it exists on TPM
I suggest you visit Bjoern Roy's web page and look at his TPM screen shots. Once you do so, it will be impossible for you to say "I don't know if it exists".
on my Pioneer Elite 710HD, it looks fantastic
Hubert, how close are you sitting to your 710HD? I sit about 10 feet from my somewhat smaller 610HD and I can see the ringing quite plainly.
As to it being stylish or popular to complain about EE or ringing or whatever... I can't speak for others but I get no joy or delight out of it. I bear no ill will towards Lucas or anybody. I just see something I don't like, something that isn't right, and I'd like to see it fixed. That's all. I'd be happy as a clam if this disc was absolutely perfect. But it isn't. There's no denying or excusing it.