It seems that my camcorder, can't archive the extreme resolution projected in the GG.

I have a Sony Digital8 camcorder, with no sharpness or edge enhacement enabled at all.

Is there any way to fix this in post?

http://www.fdivisions.com/pt_demo/pic01.jpg

Thanks in advance.

You're definately experiencing a problem--but it's not from the GG. Recording the image on the focusing screen is no different than recording an image with no attachments--it's just light entering the lens. And I have a feeling many poeple will agree with me when I say that the resolution on any 35mm adapter is going to be less than without the adapter, if at all. "Extreme" resolution on any 35mm is, I'm sad to say, wishful thinking.

If you look at the hand in the lower left corner, you can see shadowing from two other positions of where the hand has been. This suggests to be that this is an interlacing problem, even if the lines are vertical and not horizontal. I don't know how to fix it other than to say deinterlace, but that's my theory.

Life Sucks!

the vertical lines don't show up when i use "blend fields" method, for 30p.

here's a small sample : http://www.fdivisions.com/pt_demo/30p_35mm.wmv

---

I wish i could go 24p, without suffering from that problem when discarding fields.

I had a look to Your sample. There's nothing wrong except temperate chromatic abberation. But on the image it really seems to be lintarlace lines. Do You use camcorder left or rightside up?

That footage looks okay.

Roberto.

I am only guessing here but it seems to me that your camera is getting too much new information with each frame and is pixellating the image to a lower resolution to compensate.

Is your groundglass a moving groundglass. If it is, try reducing the camcorder shutter speed to 1/50th or 1/60th of a second. It is a bit hard to tell, but it seems to me from the .jpg image that there is a strong grain effect from your groundglass. If this is moving and the camcorder shutter speed is high then there's too much information for the camera's processor to cope with.

Another issue may be that you have passed from the maximum optical zoom range into digital zoom when you have zoomed in to get close to the groundglass. If you are going into digital zoom, then the image suffers such like what I see on the .jpg.

If this is what is happening, you need a close-up lens so you don't have to zoom in so much to get the groundglass image framed.

Another cause which aggravates grain effect from the groundglass is having a lens which is too close to f5.6 aperture or even tighter than f5.6. F1.8 lenses are best, wider even better, f2.8 okay if they are good, f3.5 lenses are starting to get a bit tight. f4 lenses are going to cause a problem in high contrast outdoors lighting where at least one extreme of the contrast will flicker as a grain effect or swirl or whatever. This will add to the workload of the data processor in the camera.

I am really only guessing here.

Is your groundglass a moving groundglass?
Yes it is.

try reducing the camcorder shutter speed to 1/50th or 1/60th of a second
is at 1/60th all the time.

there is a strong grain effect from your groundglass
No big deal, compared with the stairstepping on the edge of the subjects.

Another issue may be that you have passed from the maximum optical zoom range into digital zoom
NO.

Another cause which aggravates grain effect from the groundglass is having a lens which is too close to f5.6 aperture or even tighter than f5.6. F1.8 lenses are best, wider even better, f2.8 okay if they are good, f3.5 lenses are starting to get a bit tight. f4 lenses are going to cause a problem in high contrast outdoors lighting where at least one extreme of the contrast will flicker as a grain effect or swirl or whatever. This will add to the workload of the data processor in the camera.


Here's what i have:

- Camcorder ----- Sony DCR-TRV110 ( f=3.6 - 72mm / 1:1.4 @37 )
- Lens ----------- Minolta SR-7 "50mm" ( Auto Rokkor-PF, 1:1.8, f=55mm )


-------------------------------------------------------------------------


This is my RIG:

Cam w/ Little Zoom -----> Macro(0.5x) -----> GG (1.5cm from MACRO) -----> SLR Lens

That looks like bad motion blur, blown up to around 2x. I stepped frame-by-frame through the WMV, and I saw nothing that looked that bad, and nothing that isn't just plain motion blur.

I highly doubt WMV compression could have gotten rid of that kind of quality. I think your video editing software, or whatever you used to take that snapshot, was just not quite dealing with the motion blur properly. Maybe try moving the camera slower, not whipping it around.

That looks like bad motion blur, blown up to around 2x. I stepped frame-by-frame through the WMV, and I saw nothing that looked that bad, and nothing that isn't just plain motion blur.
There's nothing wrong with the video, that's just a 30p sample with "blended fleds" mode.

The picture from the first post, is an example of how the image looks when one of the fields is discarded.

Why would i want to discard fields?.
easy. look around for all the 60i to 24p conversion methods.

I think there's a flaw in your logic there somewhere. Discard half of the image, and you're just asking for the software to give you artifacts.

I wish you had said that the problem is based upon converting your 60i footage to 24p. The way your post was worded, it said nothing about anything relating to 24p, it seemed to more talk about the idea of your camera not recording video properly, instead of the 24p conversion. I automatically assumed that the screenshot and the video came from the same video, and thus, that caused the miscommunication.

I think that at this point, the problem is the 24p conversion, but I guess you had figured that out already.

I would now concur with Tom, Roberto and Ben in their assessment of your problem. With the combination and settings you have, there must be some other cause.

Bob, forget about what i said about conversion... ( i posted a video sample, JUST to show how dramatic is the difference if the fields are blended )

I'm not doing any conversions at all, PERIOD.



Check my answers in post #8, and tell me what you think.

The picture from the first post, is an example of how the image looks when one of the fields is discarded.

This is what's throwing people off. If you're taking out just the odd fields, or just the even fields, your footage will have the artifacts you've described, period. That's ordinary.

So, to clear things up: In regards to the first post, showing the picture of a frame from your footage, are we looking at complete, interlaced raw footage from the camera, or just one set of fields?

Judging by the dimensions of the framegrab, you're using your camera at a 90 degree tilt. Which would explain the interlace lines going vertically and not horizontally.

interlaced raw footage

Roberto.

In my post I used your name when I should have used Tom. Never mind.

Okay a few more questions :-

Is your 35mm adaptor a flip version or are you flipping the image in post?

Are you capturing a digital signal via firewire or capturing via an analog input?


If you are flipping the image after capture then this may well be where things are coming apart.

Flipping the image in post brings with it some difficulties.

Simply flipping the image in your editor may work but many editors seem to have minimal facility for manipulating or adjusting the image without introducing faults or deterioration.

I tried flipping the image from an earlier Agus35 I built. While the image flipped okay in Premiere 6, there was a softening and it was necessary to de-interlace and when exporting to "movie" which creates an .avi file the field orders I think had to be reversed. It is a long time since I messed with this issue. Adjusting the contrast and grading the colours only made it worse.

I have encountered this vertical banding on another camera after I have adjusted the image. It was of a different form to yours but it was a vertical artifact and in the colours.

Premiere 6 and some other basic editors do not permit me to apply effects and adjustments to a higher quality as After-Effects did. When you adjust a highly compressed file directly, there will be more errors than if the file is uncompressed, the adjustments applied, then recompressed again on export.

If there are already some artifacts in the image, then these may become aggravated by applying adjustments to a compressed file.

I found the only satisfactory way to flip the image and do colour, contrast correction was to take the captured file into another computer which had After-Effects and use in part the method published here at dvinfo on creating the motion signature of film.

This involved importing the file, using the interpret footage function, selecting highest quality, selecting upper field first, importing a second copy of the same file, interpreting footage, this time using the lower field first.

The method described is for blending the fields and some frames I think and reproducing as 24P. In my case, I do not change the frame rate but simply use the composition in After-Effects to flip the image, letterbox it, (shot it in widescreen on a PD150) then export back out.

I have used the 25P frame rate as this is a simple division of the 50i of PAL video.

I can't recall now whether I reversed the field order on export or not. But the result from After-Effects when exported out to Microsoft DV was heaps better than Premiere 6.

I can't recall the title of the article now but it should be findable in the resources section of this site.

There are two related articles, one on replicating the motion signature of film, the other on achieving the look of film in terms of colour and saturation. The author preferred the look of Sony HDCAM so the method yields an adjustment towards the Cinealta look.

Hopefully this gives you some sort of a lead to a solution.

A quick and dirty test would be to hold the camera and adaptor upside-down and shoot a short file and see if this can be adjusted for colour and contrast but without the flip.

The soft areas in your .jpg image do show some blotches consistent with grain seen by the camera on the groundglass. How big is the area on the groundglass you are shooting. I use a frame on the groundglass of 24mm x 18mm for the PD150 4:3 frame and about 30mm x 18mm approx for the FX1.

If you are zoomed in onto a smaller frame on the groundglass, then individual specks on the groundglass are going to be larger in respect to the size of the image and may cause greater workload for the processing in the camera.

Your groundglass?. What grade of finish do you have. ie., 300 grit, 600 grit, aluminium oxide of 5 micron grade?? Tell us a bit more about this and there may be some other clues to be had.

Is your 35mm adaptor a flip version or are you flipping the image in post?
I'm using my camera upside down. The image is recorded normally.

Are you capturing a digital signal via firewire or capturing via an analog input?
Firewire

Your groundglass?. What grade of finish do you have. ie., 300 grit, 600 grit, aluminium oxide of 5 micron grade?
I'm using the frosted CD from the Maxell 25-pack

I'm recording a 16x10mm area. I have have to go this close, otherwise i suffer from the black corners.


Here's what i have:

Sony DCR-TRV110 ( f=3.6 - 72mm / 1:1.4 @37 )
Minolta SR-7 "50mm" ( Auto Rokkor-PF, 1:1.8, f=55mm )


My rig is very simple:

Cam --- > (0.5x Macro)1.5cm away from -----> Frosted CD -----> SLR Lens





------------



I tried also, a lot of DV codecs, and the only one that gived me better results, was the latest version from MainConcept.

Using Canopus DV codec + ReInterpolate411 plugin, was a lot better, with amazing quality, but the vertical banding was still there.

I'm using the frosted CD from the Maxell 25-pack. >>>>>>>>>>>>>> If you hold this up to a strongly lit scene or in front of a lamp and look through it by eye, can you see things through it?

I'm recording a 16x10mm area. I have have to go this close, otherwise i suffer from the black corners.>>>>>>>>>>>>>>> This small an image is not going to be very sharp. I am surprised a f1.8 lens vignettes or has severe edge brightness falloff as you decribe. Silly question perhpas, but is this lens aperture set to wide-open. If it is not controllable manually, there may be a small lever in back of the lens which works the aperture. If te aperture is wide-open, there should be a big window of light coming through the back of the lens whe you look at it directly from behind. If there is only a very small window, then the aperture is likely closed to whatever the smallest setting is available in this lens for exampkle someting like f11 or f16 or f22 even.

If this is what is happening then if your camera is attempting to brighten up a low light situation, then the vertical banding could be a result from that. I get vertical bands from the JVC KYF50 when the light is very low and cam is on full gain.


Here's what i have:

Sony DCR-TRV110 ( f=3.6 - 72mm / 1:1.4 @37 )
Minolta SR-7 "50mm" ( Auto Rokkor-PF, 1:1.8, f=55mm )


My rig is very simple:

Cam --- > (0.5x Macro)1.5cm away from -----> Frosted CD -----> SLR Lens





------------

The 16x10mm area being recorded is the hotspot. The vignetting is just the edge of it. I'm not a pro but this sounds like a job for a condenser lens. Good Luck.