I agree, personally I have had far better results opening up the camera than trying to correct in post. XHA1.

I often shoot some night shots in very low light on a fig rig in the crowd where it’s impossible to bring in big lights
So when I edit that clip I duplicate track and use add as compositing mode, it’s amazing how the clip pumps up. Use the sliders if the highlight washes out

Rory! Excellent!!!

"[b]So when I edit that clip I duplicate track and use add as compositing mode, it’s amazing how the clip pumps up. Use the sliders if the highlight washes out [b]"

Too obvious, thank you.

Levels anybody?

Grazie

Serena is right on the money. Also some cameras do use dynamic noise reduction. Much better to have that done before the video hits the HDV encoder.
For stage shows where lighting can be all over the place general trick that seems to work best is to set exposure for the brightest lighting and don't change it. If the stage goes dark the video goes dark. Blue gels are the worst and the modern lights that have remote control gels make it all too easy for lighting to change color many times very quickly. Try to chase exposure and you'll make a mess of it. I've got some very ugly shots thanks to deep blue gels.

Bob.

Thanks everyone.

xfx -- that's thinking outside the square. Am doing something similar with a musical theatre audio (sound off the board still very low -->> 6 dups) but never thought of applying to video.

Yes, my experience is telling that in camera is best. I avoided that because of the inevitable blow-outs, but i should have gone with it. sigh.

Re levels -- levels/gain seem to have the effect of washing out the image fairly quickly. But *curves* have proven to be a great boon.

Of course, time starts to come into the equation as well. A lot of grading/correction impacts edit time and render time.

Next time, in camera it is.

Thanks again

marks

Oh, certainly in-camera is THE way to go. Sure. When you can . .when you can . . If you can't, and your rep depends on it, then walk away and ONLY do the work that you know you can cover in-camera - plus lights(!).

Unfortunately not everybody has the largess or budget in what they do do, to earn a living.

However, at least "knowing" the difference is the call we all can make, and thence from a position of awareness.

Grazie

The problem is you often edit someone else’s footage if you have a snag “adding” helps pumping under lit clips

In general some guys are strong on filming and week on editing or vice versa to compound it there’s the audio animal and then that technical monster

which gives me a bit of a byte in the backside Bus and at this bit rate makes me chipset leaving me lossless

So being effective in all 4 areas takes some doing
One thing I can say is on this forum you get a good advice

It seems that I get the same results using the "add" compositing mode as using the levels-specifically input end.

Is there advantages/disadvantages to using one method over the other?

thanks
kairosmatt

Sorry Kairosmatt you lost me there, explain in more detail

thanks

Since I haven't tested this method of adding tracks, I'll put this as a question. The first post on this thread was concerned with getting the best signal to noise under inadequate lighting. There are various techniques in post for fixing overall under-exposure (levels and curves, mostly). The technique now suggested is to take a copy of the clip and add that to the original, so doubling all levels (including noise). How is this better or different to simple amplification using levels? Simple adding (or adjustment of levels) cannot change S/N, which is the fundamental problem of under-exposure. You can do good things by compositing portions of the image, where you might separate the main subject for treatment relative to background (so keeping a lot of distracting noise under control).
In other fields of work S/N is greatly improved by stacking successive exposures, because the noise is random and the subject constant, but this isn't the suggested process .

I think what he's questioning is the value of the "adding" technique.

I too have some problem with this concept. Adding a frame to itself is no different to multiplying all pixels by 2. Everything in the frame gets brighter including the noise. With a carefully crafted curve one can avoid adding gain to the bottom of the lowlights where the noise is worst.

Now there is a technique used in astronomy that works exceedingly well. It involves adding many different frames (possibly 1,000s) of the same thing. This works because the noise is random and hence cancels out. The wanted image is not random and therefore adds up. The complication here is tracking a moving object and aligning all the frames.

Bob.

[edit] Serena you beat me to it by a parsec :)

Since the tracks that you're stacking are absolutely identical, there will be no diffference between adding the tracks or simply bumping up the gain using the levels filter.

Ok I am with you

In front of me on the timeline I have an instance of some people under a neon sign

1. “adding” the clip will allow me to control the light and the detail “noise” in the light and in the subject, levels fx does not have the same feel
2. “screening” the clip will flatten the blacks so your mid tones become more pleasant, levels hasn’t the same feel

As a youngster at art school my goal was to paint as realistically as possible that impressed me but as I matured I peered into the same paintings
What impressed me before was actually quite dead so I suppose for me the emotional content is more important so that’s what I look for
Is the clip more expressive this way then that’s what goes

Technically I am as rough as a warthogs knee so I really appreciate all your comments and am taking them to heart

Thanks as usual

Rory

Yeah, Roars! I'm with you chummie : "What impressed me before was actually quite dead so I suppose for me the emotional content is more important so that’s what I look for Is the clip more expressive this way then that’s what goes

Take absolutely NO heed to these technocrats!! They tend to take things far too literally. Emotions . . . that's what drives science! - Bottom line, they know it too, but just ain't gotta around to saying it - yet. I'm still working on 'em . . . Bob's nearly there . .. . lol . . ..

( ooo I can hear the keyboards going a-clatter as I post this . .. )

Grazie

Emotions drive science. Interesting view. Sounds like you're into creation science.

"“screening” the clip will flatten the blacks so your mid tones become more pleasant, levels hasn’t the same feel

Levels will NOT have the same feel indeed. It is a very blunt instrument. Curves are what you need to investigate. They are more organic.
Using different compositing modes will alter the maths being used dramatically, you are no longer doing a simple add.

Bob.

Curves are what you need to investigate. They are more organic.

he he he . . . that's what I've found too!!! And I've done much investigating in my time too .. .

Grazie

he he he

I know the funny cigs are popular with artistes. However, that aside, curves must be the first tool for any significant image adjustment. Nothing new or startling to be seen there.

This is a very interesting thread...'cause I do have some dark noisy stuff that I would like to improve, even if its just a little.

I have been using the curves, but I am totally just making it up. Gotta love that 'restore' option. I'm still learning the 'science', but its dead easy to see when it looks even worse!

Cheers
kairosmatt

I read most of the posts....

I'm going against the grain a little here, as I feel you should not gain up the camera unless certainly needed, if you know you're going to post you're footage for treatment. NLE's don't add noise, but can certainly bring it out or hide it some.

The image sensors quality will mostly dictate noise, but other factors affect noise also....

Gain in camera followed by Neat Video, the denoising plugin, in post. Beware though that the interface between it and Vegas dictates that the video vets delayed with respect to the audio by as many frames as you set as "radius" in that denoiser. Well worth it though. A nuisance but can most convenently fix by employing - guess what - audio delay FX in Vegas on its audio track. Just remember to match it to the number of frames (radius) in Neat Video.

Based on much experience of shooting with both Z1 and EX3 in dreadfully low-light situations, where even the (gained) EX3 looks noisy.

Sure wish Sony & NeatVideo could (jointly ?) sort out that interface matter.

"you should not gain up the camera unless certainly needed"

That's true, but the situation discussed needs more gain. It is a matter of S/N in the final image. Hiding noise in post can be relatively involved, so the question is whether that task is made easier by increasing camera gain in shooting. The final image will still be noisy, but when technique is mastered the final result can be aesthetically satisfactory.