More concrete info on ProHD

Barry, did you read Steve Mullen's article about ProHD? Does this format still use interframe compression? Still no nice discreet frames like the Panasonic camera we're expecting?
 
When you go to the page there is a picture of a girl holding a camera on the cover of the magazine. Is that a DVX or could it be something else?

I don't know what issue that is. Maybe some photo shop guru could blow it up.
 
I remember all of the transition from 1" to D2 and D1, Digibeta etc., and I always felt like I was getting bent over. All of the equipment was north of $60K.

For some reason I'm excited to see new products from Panasonic, Sony, Ikigami, and JVC, all with significantly better quality and prices that are getting incredibly low. Maybe after all of these years of dealing with "trying to keep up with the Jones's" this is actually going to benifit the consumer.

I'm sure that the Panasonic HDX will be a competitive product, but that JVC looks pretty sweet. This is cool :thumbsup:
 
Gary_McClurg said:
When you go to the page there is a picture of a girl holding a camera on the cover of the magazine. Is that a DVX or could it be something else?

I don't know what issue that is. Maybe some photo shop guru could blow it up.
Click to expand...

That's an older issue, a year or two old, maybe more. I'm not sure what camera she's holding, but it looks like a Sony VX to me.

About the article though, in reference to the PCM audio code crunching trickery, do all HDV camcorders have to perform similar tricks with audio or is it just this camera? It smells...well, I mean it sounds fishy :)
 
ccroo said:
Barry, did you read Steve Mullen's article about ProHD? Does this format still use interframe compression? Still no nice discreet frames like the Panasonic camera we're expecting?
Click to expand...
Exactly -- no discrete frames, instead it uses a group-of-pictures of six frames. It's the same as HDV, with all the same benefits and all the same drawbacks.

However, I've got to admit I didn't really follow what Steve was trying to describe as far as how the motion is created. He made it sound like the camera's actually running at 48fps, then blending two 48fps frames together to create each 24fps frame.

I hope I read that wrong, because if that's what they're doing, it had better be optional! If that's the only way it records, I may have just lost all interest in that camera (which is too early to say, of course, until we see the footage, but I want my 24 frames shot as 24 distinct filmlike frames, not some weird blended 48fps thing!)
 
Maybe. But if it does, it seems inevitable that it'll look more like video than film does. Video (as in 60i) has perfectly smooth motion ... but it looks like video. Film has stuttery stroby motion, but it looks like film. I think those of us interested in 24p want it because we want our productions to have that "film" look. Smoothing out the motion (by shooting at 48fps) wouldn't be something I'm looking forward to. It's all speculation until we see the footage, but this has certainly given me pause.
 
Seems to me that shooting at 48fps (assuming a 1/48th shutter) and combining to 24fps would look a lot like 24fps shot with a 1/24th second shutter.

I don't think that's what the article means, but I couldn't tell what it meant if not that. Weird...
 
My 2 cents:

Given that the film market is heavily biased against video, even 24P, we need pictures that come as close as possible to our film grained 35mm friends so we can make money delivering movies.

Yes there are all sorts of arguments to break with the "film look" and embrace the video look, well, you can fight that one with the distributors and show me your deal agreement afterward. I'd be happy with a beautiful 35mm film look and distribution and the check.

We need tools that will get us in the door. For those that want to do birthdays, etc... my little Sony DV keeps me happy for this stuff. ( it's a good lil camera)

We must have a Film Look for narrative commercial work, this is the 1st priority.
1. 24fps is part of the look.
2. True progressive scanning is also a must. ( this is really the most important issue)
3. Latitude is super important and color depth too.
4. Resolution is important. 720p scans well into 35mm prints and it looks good on screen. I think 720 is the lowest acceptable resolution for production that can go to film, DVD and television.

All the other stuff I whine about is important but we could survive it in fits of moaning misery. The JVC must deliver a solid performing "Film Look."
 
I asked Steve to clarify what was meant, and he basically confirmed that yes, it is doing exactly what I thought -- shooting at 48fps, and blending the images together to create a simulation of 24p, with the goal being to reduce the "judder" of 24fps acquisition.

Holy cow, I hope that's wrong.

That would totally ruin the whole point of shooting in 24P! One shoots in 24P because they're trying to emulate the look of motion picture film, which runs through the camera at 24 discrete frames per second. The judder is an integral part of the process! If they "smooth it out", they'll likely destroy that magical quality. Heck, if I wanted smoother motion, I'd shoot 60i -- that gives perfectly smooth motion with no judder effect at all. It's the judder that turns 24P into what it is, and what we like so much.

So, well, I'm hoping that the way JVC is implementing it will make it an optional filter. If they make it such that it always records its 24p that way (which wouldn't be 24p at all, it'd be something more like CineFrame 24 at that point) then I can safely say I just lost all interest in this camera. I mean, I'll hedge by saying "let's see the footage first", but I can't fathom how this would be a good thing.
 
It is a kind of imitation of the "new" 48fps filmcamera. The JVC records 48p and "blends" 2 frames to 1 as I understand it. It should give a nice smoother movement. The 24fps "magical quality" is also a limitation. The judder effect is something most of the DPs don't want to see. I guess the Pal JVC would be a 50p->25p camcorder and the final resolution would be higher than the Z1's. It looks nice to me.
 
Well, I do not believe in all this "The judder effect on a video camera, that records in 24p, is the same judder effect of a film camera". I watch so many films in the movie theater, and there is nearly no strobing in (quick or slow) panning shots. So I am looking forward to the new JVC invention of putting two real progressive frames in one (as it is described in the article) in a very hopeful way.
 
As to whether or not 24p strobes -- renders motion -- exactly like 24fps film, yes, it does. You will see quick pans in movies in the theater, but they will either be so quick that all imagnes are blurred, or they will be following something which is the center of attention -- thereby distracting you from the stroby background. You'll see it if you look. These are all techniques which have been developed over many decades of film shooting, and they work. Treat the 24p cam the same way, and you'll get the same (motion render) result.

TWO 48fps frames combined into ONE 24fps frame will NOT be the same as 24p. There will be a slight difference in any motion between the two -- there HAS to be. That difference will have to be reconciled, and there will be a resultant loss of quality.
 
Actually, I probably would not mind having this blended frames filter. Although I agree with you Barry, it should be optional (menu selectable). But for a lot of work that I do, it would be very interesting, if this filter is well done. I don't think it will emulate motion to the extent of 60i, but having something in between 24p and 60i sounds great. But I won't hold my breath until I see the results. If this filter is improperly done, it could look really bad. But I'm guessing that this "feature" will fill in the gap for those that want 24p, but are a bit nervous about the perceived 24p judder (which is very subjective).
 
Jubal28 said:
TWO 48fps frames combined into ONE 24fps frame will NOT be the same as 24p. There will be a slight difference in any motion between the two -- there HAS to be. That difference will have to be reconciled, and there will be a resultant loss of quality.
Click to expand...
You mean because of the added pseudo motion blur the filter would create? I think you're right about that. Nevertheless, I'm curious how it will be rendered. Only time will tell.
 
I do not know for certain if it is possible to mimic the attributes of film stock running through a cine camera without two ingredients that cine cameras have that video cameras do not:

a real shutter

accurate exposure properties, for lack of a better term



shutter is what determines motion. 180 degrees (video term 1/48 at "24" frames) would be considered pretty standard.

a 45 degree shutter (1/200,apx. I don't have the calculater in front of me. at "24" frames) is referred to as a narrow shutter angle and is a primary contributory to the stroby-motion look of fast-moving objects or just about an handheld camera movement. See the opening of saving private ryan for an example of that look.

sometimes a cinematographer will shoot a narrow angle to obtain a perceived increase in sharpness, as your camera is stopping motion more often the images lack motion blur... but all you are really changing is motion attributes and not sharpness.

sometimes a cinematographer will shoot a 200 degree shutter for slightly more motion blur and perhaps some may notice it, though to me it is a negligable difference from 180. I understand many Panavision films are shot this way, but I have not been on a Panavision shoot that has to confirm this.

video cameras just don't have a real shutter. that is a blanket statement, and is directed at any-non-varicam camcorder. i have used the varicam and understand reasonably well that its' shutter broken down in to angles and divisions of a second, so for folks familar with a cine camera would be comfortable with it and those familiar with video would be comfortable as well. HOWEVER, I am not entirely convinced that todays electronic implentation of a shutter angle actually presents accurate approximation of how that shutter would function in a cine camera.

For example, on the varicam the method of changing frame rates is using the "Synchro-Scan" shutter. So, the same place you dial in your 180 degree shutter you keep scrolling until you hit settings such as 1/48 or which creates 48 frames per second images and so on. A cine camera simply does not work that way; adjusting your shutter doesn't change your frame rate... especially contrast this the DVX where "Synchro-Scan" is a shutter-only adjustment, right Barry? That is, until you start doing the "12fps and 15fps" shooting methods, right Barry? I would love clarification on this, honestly if I am incorrect on any of this, but so far I understand right now as I am typing this post, none of these cameras have an accurate electronic implementation of a shutter. Shutter angle or shutter speed simply should not affect frame rates -or- the illusion of frame rates. I think the reason why these little cameras look stroby is due to the fact there is not a proper shutter on them to allow for anything else.



OKAY, my thoughts on exposure controls.

The weirdness of video (to me) is that if I overexpose an image all that means is that my whites are clipped and my blacks are compressed and aren't really black anymore and my colors are less saturated.

On a negative film, single stop of over exposure does clip whites, but it enriches blacks and over saturates colors since you are in effect allowing more light to hit your film resulting in more saturated colors. This all assumes you go for a "one stop pull" in telecine and not going to print film.

The converse is true of underexposing a single stop on negative film, which really wasn't that good of an idea until the new stocks started handling it better... Anyways, underexpose a stop and request a "push process" on transfer and you would end up with a reduced contrast image with a mellow color palette.

Has your video camera ever acted like this in regards to whites, blacks, and color saturation? Mine hasn't.

Please correct me on how I could understand the shutter of a video "24P" camera and how manufactures intend to correct the nuances of its' implentation to assist in its' "film look."

Additionally, please correct me if I do not understand the properties of over-under exposing negative films and the processing used in telecine to optimize your choices.

If you really want to get in to the nitty gritty of why video lacks an authenthic film look, read about why Kodak films rely on a dozen or so different-sized emulsions to increase dynamic range similiar to the way our eyes have rods and cones to detect varying degrees of intensity. Very interesting stuff!! CCD's contain lots pixels but they are (to the best of my knowledge) identical in size... it is no wonder video cameras have limited dynamic range compared to our eyes! And we were born from a cosmic collision and evolved... um, right. If only video manufacturers were God, and no that is not implying that Kodak is, :-)
 
One other thing: How is it possible for an inter-frame compression format to record discrete frames (rhetorical ?) Explain to me again how JVC is implementing Progressive, discrete frame recording to tape without using an intra-frame method?

I need to learn more about this compression format before I open my mouth any further, but it seems reasonable to think you can't do it.

Brian
 
Awesome post Brian

Very strong points
Now that we have achieved desirable frame rates, I think there should be more attention brought to the shutter systems within these camcorders.
When, if at all do you think we’ll have better shutter systems implemented into these camcorders?

I’ve often speculated about possible solutions to achieving greater dynamic range and color with existing sensors that are being introduced on DSLR cameras, but I never spent much time focusing and/or speculating on improving the shutter system.

With regard to the typical camcorder sensor though, I’ve thought as I’m sure others here have about the possibility of a hybrid sensor developed with Fuji’s SR CCD design in mind, as used in the S3 Pro DSLR to achieve a greater dynamic range, and the Fovean sensor as used in the Sigma SD9/10 to overcome issues with achieving accurate color. Wouldn’t a hybrid sensor between these two be a possible, albeit costly solution?
I realize that there will be better solutions in the future, but I was just speculating on what could be done with what we know currently exists in sensor technology. As far as shutters, I haven’t a clue.
 
thisiswells said:
One other thing: How is it possible for an inter-frame compression format to record discrete frames (rhetorical ?) Explain to me again how JVC is implementing Progressive, discrete frame recording to tape without using an intra-frame method?

I need to learn more about this compression format before I open my mouth any further, but it seems reasonable to think you can't do it.
Click to expand...

Capture and compression are two different things, and there's no reason that interframe compression can't be used for progressively captured frames. All interframe compression does is save space by storing the difference of the current frame from the frame before it, and this can be done with both interlaced and progressive images.

An interlaced camera will create a series of interlaced images, and a progressive scan camera will create a series of progressive images, but the compression scheme doesn't really care about how the images were made. As far as an interframe compression scheme is concerned, the CCD hands it a series of images in each case, and it just needs to remove information that is redundant from frame to frame.

Note also that interframe and intraframe compression are not mutually exclusive; I'm fairly certain that all interframe schemes also use intraframe compression for I-frames.

Edited To Add: If you need proof that progressively scanned images can be encoded in MPEG-2, pop just about any DVD of a major motion picture into your player. DVDs of features that originated on film are almost always encoded to disc at 24p, and the DVD player adds in a 2:3:2:3 pulldown dynamically. MPEG-2 is the interframe compression scheme used by both ProHD and DVDs.
 
You must log in or register to reply here.
Back
Top