Jump to content


Photo

Super-8 500T question about noise in the white parts of the scene


  • Please log in to reply
19 replies to this topic

#1 Alessandro Machi

Alessandro Machi
  • Sustaining Members
  • 3318 posts
  • Other
  • California

Posted 24 January 2008 - 11:36 PM

I find the overall level of grain for the Super-8 500T to be quite acceptable. However, the white parts of the scene seem to exhibit excessive grain, especially when compared to other parts of the scene. If I avoid white in my shots, the scenes are really very acceptable from a grain point of view, and we're talking Vision2.

By excessive grain it almost looks as if besides the grain I can see some electronic buzzing going around. If it were only grain that would be Ok with me.

The "overexpose by one f-stop" rule for negative film stocks might need to be modified for super-8 negative transfer work. It seems that if one overexposes the overall average exposure of the scene by one f-stop, the white parts of the scene may actually be anywhere from 1.5 to 3.0 f-stops overexposed. I assume the excessive white grain anomaly is probably not as noticeable in 16mm and 35mm. it appears that since the white parts of certain scenes when shot in super-8 negative are so small that the transfer system cannot actually handle such a diverse contrast because it is on such a small film transfer format.

I literally have seen images in which many colors in the scene are all minimally grainy and then there is one small piece of white in the shot that is white and the grain is screaming. Even if the white is bigger in the shot, the grain is still very visible. It is almost as if more transfer resolving power is needed for contrasty negative super-8 than if one were transfering 16mm or 35mm. My theory is the actual area of film is larger in 16mm and 35mm so there are more actual pieces of grain to represent the image so the scanner can more easily discern the contrast slopes within the shot.

Besides the most obvious but not necessarily plausible solution of not using absolute white in the scene, are there any other workarounds when shooting in super-8 to minimize the 500T Vision2 negative/electronic grain that appears over the white parts of the scene? I suppose super soft non contrasty lighting would help, but part of the fun of negative is knowing you can get away with extra contrast, so that is not necessarily plausible either. Can the film transfer facility do anything to minimize grain that is basically noticeable only over the white parts of the scene?
  • 0

#2 Adrian Sierkowski

Adrian Sierkowski
  • Sustaining Members
  • 7117 posts
  • Cinematographer
  • Los Angeles, Ca

Posted 25 January 2008 - 02:49 AM

Actually, I noticed this too on 500T 7218 in Super 16mm. I had these white cabinets, they looked "white," on the image, but they were very grainy compared to everything else. Skin-tones were nicely rendered and having sat in on the color correct, I can say we didn't touch the image. Just surprised me. only thing I could think of was the lack of contrast across the white service didn't serve as a mode of distraction from the grain. i had heard that the more contrast in a scene, the less likely we are to notice grain. As this was an all white kitchen opposite a dark living room, in a wide shot, I would assume the fact that 1/2 the frame was white and the other 1/2 dark is what caused my eye, at least, to pick up on the graininess.

I guess next time I should over expose everything by 2/3rd of as stop to help minimize grain. . .
  • 0

#3 Alessandro Machi

Alessandro Machi
  • Sustaining Members
  • 3318 posts
  • Other
  • California

Posted 25 January 2008 - 09:17 PM

I guess next time I should over expose everything by 2/3rd of as stop to help minimize grain. . .


The problem is if you overexpose by 2/3rd's of a stop then the white becomes anywhere from 2-3 stops overexposed. Is it possible that whites should be underexposed a stop and darker tones be overexposed a stop? (in relation to each other)
  • 0

#4 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 25 January 2008 - 11:55 PM

Usually it's noise, not grain, you are seeing in the white areas of the frame, if this is a negative-to-video transfer. It's particularly likely if the negative is very dense (overexposed.) Some telecines are better at handling dense negatives than others.
  • 0

#5 Alessandro Machi

Alessandro Machi
  • Sustaining Members
  • 3318 posts
  • Other
  • California

Posted 26 January 2008 - 10:00 AM

Usually it's noise, not grain, you are seeing in the white areas of the frame, if this is a negative-to-video transfer. It's particularly likely if the negative is very dense (overexposed.) Some telecines are better at handling dense negatives than others.


Interesting. So it may be rather bizarre but the smallness of the super-8 film frame combined with the higher ASA negative stocks might actually create a situation where the resolving power needed to eliminate electronically generated "noise" requires a higher sampling rate?

Is there a spec number that is attached to the resolving power of a film transfer room so one can better understand the room's ability to transfer film? I understand that there is 2k and 4k but what about when going to standard definition formats such as betacam sp, digital betacam or mini-dv?
  • 0

#6 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 26 January 2008 - 01:08 PM

Scanning at higher resolutions isn't necessarily the solution, although a scanner may generate less noise than a telecine. It's just that a very dense (dark) area of the negative requires the video signal to be pushed more, so you get noise. There are ways to minimize this in some telecines, but I don't know the particulars. There may be a way of reducing the "gain" level of the system, for example. Or apply noise reduction.

It's just a problem that people used to transferring positives to video haven't had to deal with before.
  • 0

#7 Anthony Schilling

Anthony Schilling
  • Basic Members
  • PipPipPipPip
  • 1053 posts
  • Camera Operator
  • Portland, OR

Posted 26 January 2008 - 01:22 PM

Sounds like you got a Rank transfer. You will not see that on a higher end scanner like the Shadow.
  • 0

#8 Alessandro Machi

Alessandro Machi
  • Sustaining Members
  • 3318 posts
  • Other
  • California

Posted 26 January 2008 - 05:31 PM

It's just a problem that people used to transferring positives to video haven't had to deal with before.


Ektachrome 100D was easily transferred with none of the same issues. Can I assume that Ektachrome, being a reversal film, ultimately has less individual grains to deal with and the rank can handle it better? Perhaps the contrast value between side by side grains is significantly less on reversal films than Negative films so the ranks can handle them better?

One test I came up with was to pause an actual Vision 200T negative film frame on the rank and observe if the "noise" was still there. In theory, if one frame of film is put in pause on the film transfer system the non grain noise should not be visible. If "noise" is visible, then there is room for improvement. When this test was done with the Ektachrome 100D, there was basically no "noise", when the test was repeated of the same type of imagery with the Vision 200T, there was noise.

I believe it is possible to have a rank work perfectly fine when scanning 16mm 500T or 200T but when transferring Super-8 500T or 200T noise in the white areas might be visible. By the way, the Ektachrome 100D is a real nice stock.
  • 0

#9 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 26 January 2008 - 06:07 PM

Ektachrome 100D was easily transferred with none of the same issues.


As I said, positive film doesn't have the same problem as negative film does. Ektachrome is a positive film (the image is already positive). So the brightest areas in the frame are the clearest on the film, not the densist, so there is no noise problems trying to get enough video signal from those areas in the frame.

Grain is a different issue than noise. Ektachrome looks less grainy than 500T because it IS less grainy -- it's much slower in speed for one thing. If you shot 50D negative in Super-8, you'd probably find the grain similar to 100D reversal except you'd have to watch out for noise in the bright highlights in some telecine transfers.

The problem with noisy highlights in video transfers is a reoccurring one that seems to vary a lot between telecine to telecine, post house to post house.
  • 0

#10 Alessandro Machi

Alessandro Machi
  • Sustaining Members
  • 3318 posts
  • Other
  • California

Posted 26 January 2008 - 07:55 PM

As I said, positive film doesn't have the same problem as negative film does.


I'm not disagreeing with that statement, I just want to understand if negative has more resolving power than reversal and because of that it actually makes it tougher to scan the additional contrasting grains found in negative.

Ektachrome is a positive film (the image is already positive). So the brightest areas in the frame are the clearest on the film, not the densist, so there is no noise problems trying to get enough video signal from those areas in the frame.

Grain is a different issue than noise. Ektachrome looks less grainy than 500T because it IS less grainy -- it's much slower in speed for one thing. If you shot 50D negative in Super-8, you'd probably find the grain similar to 100D reversal except you'd have to watch out for noise in the bright highlights in some telecine transfers.

The problem with noisy highlights in video transfers is a reoccurring one that seems to vary a lot between telecine to telecine, post house to post house.


If the rank was able to actually differentiate every individual piece of grain within the film frame would that eliminate some of the additional noise issue? Is it also possible that reversal film in general has less contrast differentiation between individual film grains than negative does?
  • 0

#11 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 27 January 2008 - 02:55 AM

If the rank was able to actually differentiate every individual piece of grain within the film frame would that eliminate some of the additional noise issue?


I don't know how many times I can say that noise is a different thing than grain. It has nothing to do with grain structure or contrast or resolution, it has to do with signal strength. It's an electronic phenomenon, not a film phenomenon. If something is dark and you boost the signal, you get noise, just like when something is low in volume in audio and you boost it. And on a negative, the bright areas are the darkest. This is true on slow or fast negative film.

Why would being able to differentiate between grain particles make the densist parts of the frame any less dense to the pick-up device?

The grain structure of negative and reversal are different, the graininess is different, the graininess of different speed stocks is different, but this has nothing to do with noise.

Besides, in theory, a higher resolution scan would make the true level of graininess more apparent. Lower resolution scans tend to blur some grains. But again, this has nothing to do with noise.

Trouble is that a film frame transferred to video may have both grain and noise in the image, and though they are very different things, it can be hard to tell the difference, except that grain is more "swirly" and random whereas noise is more "shimmery" and even in its pattern rate.
  • 0

#12 Alessandro Machi

Alessandro Machi
  • Sustaining Members
  • 3318 posts
  • Other
  • California

Posted 27 January 2008 - 08:18 AM

I don't know how many times I can say that noise is a different thing than grain.....


I'm not refuting this point.

.......Trouble is that a film frame transferred to video may have both grain and noise in the image, and though they are very different things, it can be hard to tell the difference, except that grain is more "swirly" and random whereas noise is more "shimmery" and even in its pattern rate.


I would suggest that digital compression cannot account for every grain in the film frame because digital, although a potentially more efficient capture format than analog video and film, has not been advanced enough to capture everything that is on the film frame, especially when the film frame is tiny.

Isn't that why effectively scanning a frame of super-8 is not as easy to do as scanning a frame of 35mm? If the rank operator pauses a frame of super-8 negative on the rank and I can see video noise being added in, certainly it would be safe to say that Rank has room to grow in their scanning capabilities. I am actually extremely happy with most Rank Transfers I get, the price/quality points are probably one of the best deals going. It's this one achilles heel of negative white grain/noise that I wish could be R & D'd a bit and improved.

Even if the density of the film negative is the culprit, shouldn't rank be able to do a more effective job if the transfer speed is slowed down?

In the meantime then, I would suggest not using white or anything near white if one wants the best looking image for 200T or 500T negative super-8 shooting.
  • 0

#13 Anthony Schilling

Anthony Schilling
  • Basic Members
  • PipPipPipPip
  • 1053 posts
  • Camera Operator
  • Portland, OR

Posted 27 January 2008 - 12:37 PM

I have noticed this on 16mm and S8 Rank transfers of 100D reversal and 50D neg. The level of graininess on the 2 stocks is very similar, but the 50D shows more digital noise. I guess I know why now. The higher end machines like Shadow or Vialta show a lot less noise. I think that is why I have been a lot more impressed with 7218 in S8 than most others... The shadow gives cleaner results, but you will pay more for it too. Maybe I will spend my George Bush economy relief check on a 7201 Shadow transfer-
  • 0

#14 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 27 January 2008 - 12:55 PM

Even if the density of the film negative is the culprit, shouldn't rank be able to do a more effective job if the transfer speed is slowed down?


This is where my knowledge of telecines runs out -- if they are anything like a video camera, then the way to reduce noise would be to reduce gain in the system, to make it essentially slower in sensitivity (not transfer speed, it should still be able to run at 24 fps or whatever, we're not talking about a huge reduction in sensitivity.) But I don't know if that is something "tweekable" in a Rank or Spirit, etc., the overall gain or lack of it in the chain.
  • 0

#15 Alessandro Machi

Alessandro Machi
  • Sustaining Members
  • 3318 posts
  • Other
  • California

Posted 27 January 2008 - 01:48 PM

This is where my knowledge of telecines runs out -- if they are anything like a video camera, then the way to reduce noise would be to reduce gain in the system, to make it essentially slower in sensitivity (not transfer speed, it should still be able to run at 24 fps or whatever, we're not talking about a huge reduction in sensitivity.) But I don't know if that is something "tweekable" in a Rank or Spirit, etc., the overall gain or lack of it in the chain.


If the sensitivity is reduced and in some way this affects the "depth of field", meaning the image isn't as sharp, that too might also reduce the video noise. Maybe there are tradeoffs going on, one can get a very sharp super-8 transfer but the moment the contrast in the scene is excessive or includes white the video noise becomes more noticeable.
  • 0

#16 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 27 January 2008 - 02:01 PM

If the sensitivity is reduced and in some way this affects the "depth of field", meaning the image isn't as sharp, that too might also reduce the video noise. Maybe there are tradeoffs going on, one can get a very sharp super-8 transfer but the moment the contrast in the scene is excessive or includes white the video noise becomes more noticeable.


Gain shouldn't affect sharpness, but electronic sharpening (edge enhancement) does make both grain and noise more pronounced.

Contrast, though, tends to improve the impression of sharpness (which is why modern reversal stocks look sharper than negative stocks) and helps hide some grain (grain being most visible in flat areas of midtones, which higher-contrast shots have less of.)

I don't know how the Super-8 gate works in a Rank or Spirit/Shadow, but if it is some sort of magnifier-type lens element, to allow the Super-8 frame to be enlarged to match the sensor area (maybe that is only necessary with a fixed line array CCD like in a Spirit rather than a flying-spot scanner in a Rank though) then maybe that adds some sort of artifact too, though I don't know what.

Or perhaps the image isn't being magnified pre-sensor... so there is increased noise in the system from enlarging a smaller scan to fill the normal resolution of the recording system.
  • 0

#17 Gustavo Lopez

Gustavo Lopez

    New

  • Basic Members
  • Pip
  • 2 posts
  • Student

Posted 31 January 2008 - 03:32 PM

Hello

I am new in the forum and I recently purchased a Beaulieu zm II 40008, and I am still learning how to use it correctly.

I have bought 2 visionT 500 super 8 cartridges and I want to film some images at my friends concert with them, I think the light will be low or medium, typical light at a concert. Please can you tell me any recommendation about the camera controls with this type of film to get images with an acceptable quality? Should I open the diaphgram at maximun? And what about the iso? 200 or 400? An for slow motion 70 fps sequences any recomendation with 500t and low light scenes, can it be done?

I hope you can give me some advice if you have experienced with this kodak film and the camera.


Thanks very much!!
  • 0

#18 Alessandro Machi

Alessandro Machi
  • Sustaining Members
  • 3318 posts
  • Other
  • California

Posted 01 February 2008 - 10:18 AM

Hello

I am new in the forum and I recently purchased a Beaulieu zm II 40008, and I am still learning how to use it correctly.

I have bought 2 visionT 500 super 8 cartridges and I want to film some images at my friends concert with them, I think the light will be low or medium, typical light at a concert. Please can you tell me any recommendation about the camera controls with this type of film to get images with an acceptable quality? Should I open the diaphgram at maximun? And what about the iso? 200 or 400? An for slow motion 70 fps sequences any recomendation with 500t and low light scenes, can it be done?

I hope you can give me some advice if you have experienced with this kodak film and the camera.


Thanks very much!!


I don't think you will have much luck shooting slow motion, probably not enough light. If there is enough light, it might be a situation where any of the performers performer being lit by a spotlight will come out but the surrounding area may be completely dark in slow motion. I tend to think you will be better off with the f-stop wide open, however you don't know if your camera lens is properly collimated so that it stays in focus in the wide position after you focus in the telephoto position.

Spotlights can be a blessing and a curse as they create more opportunities to shoot but it actually becomes possible to overexpose the person being hit by the spotlight while still not having enough light for the surrounding environment.

Another problem with shooting slow motion in a loud environment is you can accidentally run off way more film than you are realizing because you don't hear the film racing through the camera.

You may find that "undercranking" doubles the amount of film you shoot and gives you more chances for good exposures, but you will then have to slow down the frames per second in transfer to match the filming speed you shot at. 6, 9, and 12 frames per second are popular filming speeds for undercranking in situations where you don't have a lot of film and the light levels are low. But be careful, you may get more blurriness than you want if you are going handheld.
  • 0

#19 Anthony Schilling

Anthony Schilling
  • Basic Members
  • PipPipPipPip
  • 1053 posts
  • Camera Operator
  • Portland, OR

Posted 01 February 2008 - 02:11 PM

Hello

I am new in the forum and I recently purchased a Beaulieu zm II 40008, and I am still learning how to use it correctly.

I have bought 2 visionT 500 super 8 cartridges and I want to film some images at my friends concert with them, I think the light will be low or medium, typical light at a concert. Please can you tell me any recommendation about the camera controls with this type of film to get images with an acceptable quality? Should I open the diaphgram at maximun? And what about the iso? 200 or 400? An for slow motion 70 fps sequences any recomendation with 500t and low light scenes, can it be done?

I hope you can give me some advice if you have experienced with this kodak film and the camera.


Thanks very much!!

Yes, I have filmed dozens of live shows at small venues on S8 500T. I have mainly used the Canon 814XLS set on a 220 shutter, and still keep my iris nearly wide open. The only times I have gone into over exposure is on heavy bright white flashes, and the 500T still held up very well. On the Beaulieu 4008, I believe you are dealing with a 1/90th shutter and a max of f1.8? My advice would be to stay wide open and keep your framing on the best lit areas of the stage. If the lighting is some what decent, you will get good exposure on the subjects and have some room to crush the blacks without much noise. Keep in mind the guy running the sound board is usually running the lights too, and they have a habbit of half assing it on the lights. I usually lay into them before the show, tell em to give me the best they got on the lights and slip em $5. It's a cheap gesture that goes a long way on what can be a $400 film chain... the difference between beautiful sharp images, or an expensive spool of mush. Good luck.
  • 0

#20 Anthony Schilling

Anthony Schilling
  • Basic Members
  • PipPipPipPip
  • 1053 posts
  • Camera Operator
  • Portland, OR

Posted 01 February 2008 - 03:10 PM

Here are some samples of the same scene... 7218 16mm, Beaulieu R16, Angeneaux 12-120 zoom, f2.2 1/60th? The lighting was less than I wanted for this set, but was still able to pull an exposure. The flash is a random spotlight almost head on to the lens, it last about a second in real time.
Posted Image
Posted Image
  • 0


Tai Audio

Willys Widgets

Wooden Camera

Rig Wheels Passport

Opal

Glidecam

The Slider

FJS International, LLC

CineTape

rebotnix Technologies

Ritter Battery

CineLab

Broadcast Solutions Inc

Aerial Filmworks

Technodolly

Gamma Ray Digital Inc

Metropolis Post

Media Blackout - Custom Cables and AKS

Visual Products

Abel Cine

Paralinx LLC

Willys Widgets

Opal

rebotnix Technologies

CineTape

The Slider

Metropolis Post

Wooden Camera

Media Blackout - Custom Cables and AKS

Ritter Battery

Glidecam

Gamma Ray Digital Inc

CineLab

FJS International, LLC

Rig Wheels Passport

Visual Products

Technodolly

Tai Audio

Aerial Filmworks

Paralinx LLC

Abel Cine

Broadcast Solutions Inc