mini35 DVX100 ASA speed

[Mike Bluestone]

Does anyone have any idea as to what might be the equivalent ASA speed when using the mini35 with the DVX100 Panasonic in progressive mode.
Alternatively, does anyone know if it is faster or slower than using it with the Canon XL1S

[Barry Green]

The DVX100 with the mini35 on it was about 2 to 4 stops slower than the DVX with the anamorphic adapter attached.

To use the mini35 you have to zoom all the way in, and the process of zooming in costs you about 1.5 stops over the length of the lens. Then you lose about a stop due to the light loss from the ground glass. Then, the final loss was due to the slow still-camera lenses we were using (about f/2.4 to 2.8).

In our testing, we shot a scene using the mini35, and then shot the same scene using just the anamorphic adapter. We found we had to stop down about 4 stops to get equal exposure as compared to the mini35 shot.

Using faster cine lenses would improve the situation, obviously. If you were using an f/1.4 lens instead of our f/2.8 lens, you'd buy back two stops.

A DVX100 in progressive-scan mode, by itself, is 640 ASA. Assuming you're using a fast lens, you can ballpark the DVX/mini35 combo at about 125-160 ASA.

But this is all speculation based on a hazy recollection of what we were shooting, not the results of specific calibrated tests. Next time I get ahold of the mini35, I'll test for accurate ASA, as that'd be nice to know.

[Mike Bluestone]

Thank you for your help.

[Eric MacIver]

Can you tell me how to test the ASA? I should be using one of these in about 1-2 weeks with Zeiss T1.3 Primes and I can get a value for you if it isn't too complicated.

I'll be able to get it for the XL1s as well.

Thanks

[Barry Green]

Please do some 24P/DVX tests against Frame Mode/XL1 tests, using the same lens, same shot, same lighting, etc... that'd be very interesting to see! Shoot a resolution chart if you can. If you don't have one you can download one from www.bealecorner.com.

Testing the ASA is done by using a spot meter. You can use an 18% gray card, or you can get a good idea by just shooting a variety of objects. Set the camera at 1/60th exposure, and the spot meter at 1/60th exposure. Point the camera at the gray card, or at some other scene, and use auto-iris to get an f-stop reading. Then point the spot meter at the same gray card, or at the same scene, trying to fill the spot section with the same info that filled the screen on the camera. Take a spot reading. Then adjust the ASA on the spot meter until the f-stop on the spot meter reads the same as the f-stop on the camera.

Also check to see if you have any autoexposure compensation on the camera. The DVX in F5 or F6 presets has the exposure pre-set to underexpose by 3db (or half a stop). You'll want to zero that out to make sure you get accurate readings.

[Eric MacIver]

Simple enough :) I almost feel like saying "duh" on that one.

I'll be doing a comparison between the XL1s PAL (25 fps in frame mode) NTSC (30 fps in frame mode) and DVX100A (24P).

If I have enough time I may even test the XL1s's in normal mode, with Magic Bullet applied.

It'll take a few weeks, but I'll post the results.

[Dennis Hingsberg]

Eric,

Nice to know you shoot on PAL XL1s as well. How do you find your NTSC conversions? I find the results of my video converted from PAL 50i to NTSC 60i look great even if left as interlaced. I watched my film for the first time at a film festival in a theatre using a digital projector and was extremely happy with the results. It also recently screened on a smaller screen (7 foot) with different equipment and also looked great.

I have to comment that shooting on PAL and converting to NTSC after has left me with no need to really shoot progressive - or in frame mode. I get amazing motion characteristics in the transfer because PAL being 25fps is so similar to the frame rate of film's 24fps.

So let me ask, are most people shooting in frame mode or Progressive because they're going to film (35mm blow up)?

[Eric MacIver]

I haven't yet shot anything on the PAL version that has gone through the editing process. I hope to soon, and I was/am hoping to see the charactaristics you described.

I really want to do a side-by-side comparison of Frame Mode vs. Normal Mode. So, when I do, I'll sure let you know what I think :)

[Dennis Hingsberg]

I don't use frame mode, nor do I like it because the resolution is lower than in regular mode.

Good luck with your tests though and be sure to post results.

[Mike Bluestone]

Dennis
What method have you used to convert PAL to NTSC?

[Dennis Hingsberg]

There are actually several ways to have this done and I haven't tried them all (yet).

One way is to put each PAL frame into its own NTSC frame, but will result in your video playing 4% faster. This is not noticable to the viewer, but the audio must be pitch/time corrected in this method unless you have no real sync concerns then you can just lay down your audio tracks again to the video. Some adjustment would also have to be made to the frame size because PAL is 720 X 576 (768 x 576 with square pixels). The advantage of this method is that each unique NTSC frame gets a full PAL frame so there is no interpolation of fields or frames necessary. This however does not really generate a film type of look though, the final output should be deinterlaced for more of that filmic slowed down judder look.

The next method actually has several different formulas or processes but basically involves converting interlaced PAL footage to progressive video using an adaptive deinterlacer, then using that "new" footage to create a frame rate that is desired. In this case NTSC.

This last method is the one I've been using the most and some of the tools I've been using are VirtualDub, AVISynth and a deinterlacer plugin by a guy named Gunnar Thalin. These guys mainly developed these tools for cleaning or converting video formats when pirating, downloading or ripping video from other sources - but it can be used for legitimate purposes as well.

Try www.avisynth.org and www.virtualdub.org to learn more.