|
|||||||||
|
Thread Tools | Search this Thread |
April 16th, 2009, 11:59 PM | #16 |
Major Player
|
OK, I'll try again!
Really this cannot be answered by thinking about it and only through definitive tests. Broadly the question is whether it is better to sharpen at the 12 bit level of the DSP or leave it until post when working with 8 bit long GOP? And how is that changed working with 10 bit 2 frame GOP of Cineform NEO HD which has reconstructed the footage? And what are the effects of various NLE FX strings? The perception of sharpness is strongly dependent on contrast, a low contrast image appearing less detailed even though of the same resolution. At the same time, as object contrast reduces so it is more difficult to resolve fine detail in the object (effect of MTF). Assuming that such detail is still contained in the recorded signal (even if not apparent) what is the best means for recovering it? Unsharp masking? Certainly a powerful tool for this purpose in astrophotography. Or is simple edge sharpening effective? Since we're not concerned here with scientific imaging, any artefacts introduced through the DSP matter only when they become visible or effect other processes (such as chroma key). Having stuff pop into focus at the end of a pan, for example, is the reason I leave detail "off". Reviews have noted that detail -40 is about the same as detail "off", so there is plenty of margin for experiment. My view is to go with whatever settings give you the image you want up on the screen, rather than try to discover the "right" answer. However if someone can propose a definitive test then it would be interesting to explore. Last edited by Serena Steuart; April 17th, 2009 at 08:25 PM. |
April 17th, 2009, 11:53 AM | #18 | ||
Regular Crew
Join Date: Jul 2004
Location: Berlin, Germany
Posts: 169
|
Quote:
But it does not help. At very sharp corners/edges like lights in dark rooms, especially eg. candles in wind glasses, you have this very obvious oversharpened dark "aura/corona". This will only be reduced at level -20/-30 but then everything is even more blurry as without setting details to on. Everything else in the room benefits with kindly better sharpness when details are set to on. |
||
April 18th, 2009, 05:14 PM | #19 |
Trustee
Join Date: Dec 2008
Location: Kennewick, WA
Posts: 1,124
|
Came across this detail "test" that some might find interesting.
Testing Letus Extreme on Sony EX1 Notes from the Field
__________________
Sony EX3, Canon 5D MkII, Chrosziel Matte Box, Sachtler tripod, Steadicam Flyer, Mac Pro, Apple/Adobe software - 20 years as a local videographer/editor |
April 18th, 2009, 06:39 PM | #20 |
Major Player
|
The test was useful for the intended purpose but no more than the subjective assessments people have been discussing. A quantitative test is needed encompassing all the factors mentioned above if the original question is to be answered.
|
April 21st, 2009, 07:56 AM | #21 |
Major Player
Join Date: Nov 2006
Location: Brisbane, California
Posts: 530
|
So Serena, what do you personally use when you shoot with the EX? Do you vary "Detail On" and "Detail Off" depending on the subject matter?
I did a shoot the other day with Alister's setting that included CINE 1 and Detail at -8, and I was quite pleased with the sharpness. I didn't scrutinize it but to me it was noticeably sharper and I didn't feel I lost any other aspects of the image, such as the filmic quality I was getting with the "Detail Off" setting. This was a very basic outdoor shoot with some high contrast sun with shadows. Thanks all for everybody's input, if anybody has more, please feel free to contribute to this thread. |
April 21st, 2009, 06:36 PM | #22 |
Major Player
|
I shoot with detail "off" and sharpen in post where I think that benefits. But your thread has raised interesting questions about where sharpening should be performed for optimal results (DSP or post). A good question for Adam Wilt (who also shoots with detail off).
|
| ||||||
|
|