Settings for lowest noise?

J

joelnet

Well-known member
OK, I'm going to get another crack at an HVX test and I'd like to walk out of there with the cleanest footage I can. I'm looking for smooth skin tones, less saturation and lowest noise.

Can anyone advise me on the settings that will get me closest to this?

Thanks!
 
For lowest noise use black press gamma (or cinelike-V), coring +7, detail -4 or -5, and normal matrix, and for skin tones turn the skin tone dtl feature on. For lower saturation crank down the chroma level setting to taste.

And make sure there's adequate light for a decent exposure.
 
Barry? High Gamma? doesn't that boost the lows and mid, like for darker scenes?
I would think Low, BPress or Cinelike V would be cleaner although punchier.
You're doing the testing though, so I'll take your word for it.
BG EDIT: hadn't completed testing, but now that I have I will say that you were exactly right! BPress is the cleanest, CINELIKE-V is next, and LOW comes in third!

I used HD Norm for my commercial and it looked great. Cinelike D is beautiful for cinematic stuff but avoid it for Keys and when clean is a must.
 
Question on the various gammas ... it's my impression (visually) that they just punch things up differently or add a little different emphasis on colors, etc.

To keep things as CLEAN as possible, would it be best to simply stick with Normal Gamma, then CC in post to taste? Also, you'll have more latitude in post vs. being stuck with a particular look from in the camera. I know most cinematographers hate hearing (do it in post), but I'm not a camera pro yet, and my talent's in post ...

Plus, we just got done shooting our short with the DVX, and had some substantial noise ... I know it's not near the comparison of the HVX, but still the settings for lowest noise should be the same, right?
 
Normally, if it works like the DVX100A and B, the cleanest setting should be LOW or BPRESS.

HIGH gamma setting has more lattitude because it raises the lows to lower mids without compromising the highlights.

I would say "high" is pretty close to cinelike D for lattitude, but it is cleaner.

It does not show substantially more noise than the Normal gamma.

IF IT WORKS LIKE THE DVX...

"""BUT...in theory, it should.


BTW, i could adjust some of the gamma in post, but it is impossible to get the same visual texture as the "in camera" gamma curve.

I think it is preferable to shoot with settings "as close as possible" to desired look, because radical gamma correction, with any 8 bit info, will create steppy levels.

Barry ?
 
I haven't finished testing, but I thought high looked pretty good. Definitely lower noise than cine-like-D.

Low's good too. Black Press is pretty good as well; the only one that I really noticed a boost in noise on is cine-like-d.

There's too much to do, and not enough time to do it in; the detailed examinations of the settings are likely going to take place next week.

And I always definitely prefer to shoot as close as possible in-camera, always.
 
Barry_Green said:
I haven't finished testing, but I thought high looked pretty good. Definitely lower noise than cine-like-D.
Click to expand...

Thanks for the tips, I'll let you know what I find out.
 
Oh -- another thing for noise -- if you're shooting outdoors, you'll get quieter noise performance if you use the tungsten preset and a #85 filter. At least, on the DVX it made a difference, I'm pretty sure it'll make a difference on the HVX as well.
 
Okay, finished testing gamma curves for luma noise. Evin was spot-on. I'd rank them in this order, from cleanest to noisiest:

B.PRESS
CINELIKE-V
LOW
HD NORM
SD NORM
NEWS
HIGH
CINELIKE-D

I'll do chroma noise testing soon.

Did a little bit of side-by-side noise testing between HVX in DV mode, and DVX100B in DV mode. I'd say that they're almost exactly the same in terms of noise performance, when set to the same settings on each. If anything was noticeably different, it's that the HVX noise texture is bigger, wider, and smoother; the DVX100B noise was of the smaller grainier bittier texture. I think that it leads to the HVX looking a little cleaner, but it's hard to say -- the texture is just different.

And remember, the 100B was modified to have lower noise than the 100A. I'd say it's fair to say that the HVX is as clean (if not cleaner) than the 100B, and definitely noticeably cleaner than the 100A (at least when it comes to luma noise). Still have to test for chroma noise.
 
Barry_Green said:
Okay, finished testing gamma curves for luma noise. Evin was spot-on. I'd rank them in this order, from cleanest to noisiest:

B.PRESS
CINELIKE-V
LOW
HD NORM
SD NORM
NEWS
HIGH
CINELIKE-D

I'll do chroma noise testing soon.

Did a little bit of side-by-side noise testing between HVX in DV mode, and DVX100B in DV mode. I'd say that they're almost exactly the same in terms of noise performance, when set to the same settings on each. If anything was noticeably different, it's that the HVX noise texture is bigger, wider, and smoother; the DVX100B noise was of the smaller grainier bittier texture. I think that it leads to the HVX looking a little cleaner, but it's hard to say -- the texture is just different.

And remember, the 100B was modified to have lower noise than the 100A. I'd say it's fair to say that the HVX is as clean (if not cleaner) than the 100B, and definitely noticeably cleaner than the 100A (at least when it comes to luma noise). Still have to test for chroma noise.
Click to expand...

Luma noise was the main problem with the dvx100a, at least for me. This is quite reassuring! Are we to assume that if we compare a dvx100a in cinegamma d vs a Hvx with cinegamma d, we will get a noticeably cleaner picture on the HVX? If so, than I'm sold!
 
I don't have an "A" here to test. I believe that you are correct, in that the 100B had noticeably lower noise than a 100A did, and the HVX appears to be about the same noise level as the 100B. But without comparing directly, I don't know the answer for sure.
 
Thanks Barry.

Was this purely a visual test, or were you
able to use some sort of subtractive measurement?

How noticeable was noise comparing B.PRESS over CINELIKE-V?


I'm looking forward to your chroma noise tests.


Steve
 
Barry:
Is this true? The DVX is less noisy if you shoot with the tungsten preset and a # 85 filter? was your test with a plastic or glass filter? What brand? Is it a big difference? Mainly in the blacks or the entire image? Interesting...
 
Stevet said:
Was this purely a visual test, or were you
able to use some sort of subtractive measurement?
Click to expand...
Nothing too complex, just visual, using Vegas to magnify the image to something like 8x normal size to make the noise unmistakeable, and then subjective judging as to which was cleaner than which.

How noticeable was noise comparing B.PRESS over CINELIKE-V?
Click to expand...
Very close.
 
evolvnyc said:
Barry:
Is this true? The DVX is less noisy if you shoot with the tungsten preset and a # 85 filter?
Click to expand...
It is true, but as to whether it's enough for you to worry about is a subjective thing. And, let's make sure people understand -- this is only for daylight conditions, as a substitute for using the 5.6k preset.

When you shoot film, you normally pick your film stock based on the circumstances you'll be shooting -- daylight-balanced film, or tungsten-balanced film. If you shoot tungsten film in daylight, it'll turn out really, really blue... you can have the telecine correct for that by applying an "electronic 85" which will correct the colors, but it adds noise when doing so.

The video cameras appear to operate the same way. Michael Phillips of Avid first pointed out to me that CCDs are optimized for tungsten balance; compensating for daylight involves the camera doing an internal "electronic 85" to balance the signal, and that balancing can introduce a little noise. You can avoid that noise by using a glass 85 filter. However, be aware that using a filter can affect your image in other ways! A low-quality filter can harm your image a lot more than a little noise would! Plus there's a potential for ghosting or flare... you really need to get a top-notch filter, you *never* want to save money on anything that goes in front of your lens.

I believe it was Scharky who posted some DVX clips comparing the noise effect of 5.6k prst vs. 85+3.2k prst. The noise was lower, but not tremendously so. I don't know whether I'd bother chasing the filter for the noise decrease; I mention it because this thread was about how to get the lowest noise possible, and that's one way.

That said, I got a Century Tru-Pol polarizer, and it happens to be an 85 polarizer, since I figure you only typically use a polarizer outside, so -- why not whack both birds with one stone? :thumbsup:
 
Apologies to all our avian friends, and their supporters. No actual feathered beings were harmed in the writing of that previous post, nor in the acquisition of an 85 true-pol filter.

:grin: :laugh: :thumbsup: :engel017:
 
Michael Phillips of Avid first pointed out to me that CCDs are optimized for tungsten balance; compensating for daylight involves the camera doing an internal "electronic 85" to balance the signal, and that balancing can introduce a little noise. You can avoid that noise by using a glass 85 filter.
Click to expand...

This seems very logical, after all in low light we are more likely to see a tungsten balance. It would be cool if the ND 1/64 included an 85 filter. After all, the main time I'd switch it in would be when I'm in sunlight. Just a wish list idea. Now if only Panasonic could only make that as a service center mod...


Bob Diaz
 
LOL. Like I've said before..this camera reminds me of alot of motion picture camera memories. Can't remember the last time I heard "85" from video guys. Man I love HD.
 
You must log in or register to reply here.
Back
Top