24 fps editing?

Martin,

I did know of commercials shot on 25fps and screened 24fps. But never would I think that 16/35mm or HD 24p movies would be shot 25fps or "25p". One starts to think; what is this miniDV 24p camera worth. Really, the XL1 (PAL) can do the same thing in the end.

Thanks for the info Martin.

Also, excuse me for asking, but are you related to Bo F. Munthe in any way?



Regards / Med vänlig hälsning,



Erik

[Rob Lohman]

The question about the monitor is a valid one, but the bigger
DV cards might be able to handle this problem (not sure though).

On my PC monitor I have no problem whatsoever ofcourse, and
if I want to test it I can make a progressive DVD file and play that
out through my hardware mpeg decoder in my PC to a TV.

Anything yet Shawn?

[Martin Munthe]

Erik,

the benefit of a 24p DV camera would be true progressive in a prosumer camcorder. Something we have not seen so far...

And, yes, I am related to that perticular Ninja Master you mentioned :)

Rob,

on most smaller projects you are probably ok just using the CPU. On long form projects however there are great risks of sync drifts and drop frames if the CPU is the only thing working. The more referal files you have in a project the more demanding on the CPU it will be. I don't think there are any video cards for the pc that can do 24fps. As far as I know IgniterRT with Film Option is the only sub $20K card that can do it. Messing with the fps setting in Premiere could give you problems with the DV codec algorithms. Have not tried it though. It could be hazardous to the image quality output (probably producing an unsharp image). DV is sensitive to recompression as it is. I've posted some recompression examples between DV and Uncompressed D1. They are here:

http://www.operafilm.com/codecs.html

Hope it's of use to someone. Good luck with the tweaking!

[Shawn McBee]

I'm quite a way off from knowing anything, as I have to purchase several things and only have about 1200.oo a month to spend on video equipment. But I figure I'll run my idea past you fine folks and see if any of you think it's even feasable...

The idea is to get a firestore direct-to-disc interface, have a custom electronics maker come up with a circuit board that would make a brief connection between two wires attached to the chipbase 24 times each second. These wires would then be plugged into the Firestore to act as a GPI trigger.

That idea led me to think: if a circuit could be made to do that, then why not program several different "frame rates" into it and have a switch or a knob to select them. Make a little casing for the circuit board, and you'd be ready to go.

It really seems like it'd work to me, for the most part. I can see some small issues that might arise, that would have work-arounds:

1) I'm not sure how the firestore works with it's GPI trigger, but it may only record Image when the GPI is triggered, so audio would have to be recorded seperately. Or it may sound stuttered if it records sound with each tiny trigger, but not a constant sound, which would have the same solution.

2) I'm also not sure if it would record properly to the DV tape, I think it would probably either record to tape at normal NTSC (or PAL, depending) and your selected Framerate on your drive or you'd just have to forego the tape altogether and take just straight hard dirve data with no backup.

other than that, I think it could work (and on any MiniDV cam!).

So, do any of you more experienced folks think it floats, or am I just grasping at straws?

-Shawn

[Martin Munthe]

Nothing is impossible... I hope?

The thing that comes to mind with your invention is; how are you going to view this? When editing the material are you then going to send it out through the same circuitry to a broadcast monitor? Will the monitor be able to show the 24fps signal your circuit produces?

[Shawn McBee]

I hadn't really considered the monitor issue. I was mainly focused on the editing issue, which is why I started this thread.

I imagine I'd be able to run RCA out of the camera and view what I'm filming at regular NTSC from the camera before it goes to the firestore (similar to a videotap on a film camera) and use a 2:3 pulldown for editing and any video/dvd dubs, but remove the 2:3 for the possibility of a film dub.

Now I just need to learn about 2:3 pulldown... time to use that nifty "search" option!

-Shawn

[Martin Munthe]

<<<-- I was mainly focused on the editing issue-->>>

I understand that and that's why I am asking. I guess you would like to color correct your material proper to broadcast standards and with a bit of artistic flare? Thats really impossible to do if you haven't got some form of reference monitor video output from your NLE. I've tried to calibrate CRT's to do broadcast colors but it is not possible. If you are producing films for the web this will be no problem since the web is viewed on CRT's and not TV sets.

Also, if your not doing any form of CC or image adjustments I guess it would work to just cut your video stream on a CRT and then take the project somewhere else for CC. I would definately CC before I print out to film. It will cost you a lot more to CC the film.

Perhaps there is some scan converter that could do the trick?

[Shawn McBee]

I will have an actual video monitor hooked into the NLE and from what I understand if, while I'm editing, I have the 2:3 pulldown, it should be totally viewable on the monitor, so I can do any color correction or whatever else I'd like.

-Shawn

Martin,


I've also heard that the XL1 doesn't produce a real progressive image. But what is the real difference? The XL1 records 2 fields at the same time. If you put these together on 35mm film, then there will be a sharp frame even at motion. Can't the XL1 in the end be regarded as a real progressive camera although it's really not?



I suspected that you were related to the Ninja master. Munthe isn't an ordinary name, and few of them work with film.



Regards,



Erik

[Casey Visco]

Erik, the XL1 does not record 2 distinct fields at the same time....it only records 1 field through the lens and interpolates the other. And even still, these are recorded to the tape in an interlaced form. It can therefore not be regarded as a progressive scan camera.

[Rob Lohman]

Martin,

Editing 24 fps DV is no problem, also not for the DV codec because
DV is like a JPEG compression, it is not like an MPEG compression
where frames interact with each-other. So this is totally safe.
As I said earlier, the monitor might be a problem. In regards to
your statement that it is a lot of work for the CPU to keep
video and audio in sync, I have to disagree. I, personally, have
never had a problem with audio sync. If you have a pretty
new CPU (at least 1 GHz) I doubt it will have much to do. These
CPU's are very fast!

Casey,

It is correct that the camera is interpolating more or less, but
it is not as simple as capturing one field and interpolating to
get the second field. There is done more. Also the material
on tape is not layed down as interlaced footage. It is recorded
as a single frame. An NLE does not know if the footage is
interlaced or not, you usually have to tell them or they guess
it on their own. For example, the Microsoft AVI standard does
not even have a way to indicate if the source if interlaced or
not! (weird enough, but actually true).

Shawn,

I don't think I'm understanding your solution 100%, but I got
a feeling this might not work. The camera outputs a 25 or
29.97 fps signal and the firestore just dumps this signal to a
harddisk (basically). The firestor doesn't do any "capture",
so will therefor not be able to capture in intervals (that is
basically what you are suggesting, right?). This would be
the same as to do drop some frames from the actual DV stream
once you have it in a PC (only less smooth, I agree). It probably
will be more interesting to use the analog outputs of the camera,
hook it up to a professional capture board at let that thing
capture at 24 fps.

Just some thoughts....

[Shawn McBee]

Rob,

The firestore has a "REC Still" button. This can be used in one of two ways: You can press the button, record one still frame as it's own file, press it again and record another still frame as another seperate file....etc. OR... you can press the button and each time you do so it will record another single frame onto the same file. (A neat trick if you want to do a DV stop-motion movie without killing your camera's heads). The GPI port affects this feature (I think they had time-lapse recording in mind) so, really, if my idea worked, you'd be recording 24 still images per second (just like film) as one file that, when put in an NLE and played at 24 fps would, hopefully, play back at a true 24 fps, with a more filmic look. (Of course bettered by good lighting, depth of field and all that other good stuff....Just a disclaimer to keep the different film-look schools from debating over this small issue here)

I hope that explains it a bit better. If anyone knows for a fact that it wouldn't work....let me know. It'll at least save me some money.

-Shawn

[Rob Lohman]

Ah... I didn't know that Shawn, my bad. This does sound
interesting indeed. I am wondering how they do this.