View Full Version : MPEG-4 Part 10 HDVPRO/HDVPRO 70 HVR-P10 1080P60


Jack Zhang
May 10th, 2005, 12:34 AM
My first and most wildest post!

I have an idea for an MPEG-4 Part 10 (H.264) HDVPRO camcorder

The specs:

*records full 24P, 30P, and 60P!!!

*records 16 bits per channel of VIDEO!!! (total 48bits)

*audiowise, records a max of 4 channels in 24bit/96khz in uncompressed PCM!

*overall video bitrate of 50Mbps in medium-mild compression

*Square pixel recording

*a HDVPRO 70 mode to record in the 2.35:1 aspect ratio (quality not compromised: 2430x1080)

*a 120P high speed mode for 60P slow-mo effects

*Manual gamma control (a must have for pros!)

*records 30 minutes on miniDV and DVCAM and maybe DVCPRO and 90 minutes on HDCAM SR M-sized tape

*3 widescreen 8-megapixel CCDs for even more detail (6-megapixels effective in HDVPRO and the entire length of the CCD in HDVPRO 70)

*Prosumer model will include a Compact Flash/Microdrive slot (HDR-PH1000)

*Expansion Slot for BNC and HD-SDI and Wi-Fi Antenna for transfer to Gateway (512mbps internet)

*24bit DXP for faster exposure compensation

*MOTOR OFF mode for high speed ring manual zoom (mechanical mechanism: lifts zoom motor off zoom ring)

*ISO Simulator for simulated film grain

*Claims to be artifact-free and GOP-free!

*TIMECODE KEEP when Transcoding

*20x Zoom

*Shutter and F-stop Ring on lens

all of the above is possible by a IBM 64bit processor

Model #HVR-P10

--------

Impossible? or Plausible?

Let the pros such as Douglas Spotted Eagle decide.

I'd love to hear a reply!

Jack Zhang
May 12th, 2005, 09:48 PM
*Uses 4:4:4 color sampling

*Timecode and BATT REM. LCD display on cassete lid

*4 channel audio monitor LCD behind main LCD

VTR: HVR-M20

*Component in

*HD-SDI in and out

*1 DVI in/out

------

Can someone research about MPEG-4 Part 10 at 50Mbps at 4:4:4 color samples and at 16 bits per color and compressed? I'm not sure how much compression (high, medium, or mild) is needed for optimum results.

Jack Zhang
May 19th, 2005, 06:26 PM
Hello? anyone's professional opinion on this? Improvements? Rethinks? Bad features?
Doug S. Eagle? Chris H.? David N.?
I'm desparate for an answer! This is something stuck in my head for the last 1 and 1/2 years and I want to let it go!

P.S: Mods, move this thread to the HDV forum if no one answers within 4 days.

Steven White
May 19th, 2005, 07:06 PM
Well, I was wondering:

Is the combination of 48-bit colour and 4:4:4 even POSSIBLE in MPEG-4? Largely a delivery codec, I'd have expected MPEG-4 to be predominantly 4:2:2 or 4:2:0... but I don't know how well it scales up. Furthermore, why 50 Mbps? What leads you to think that's enough. I'd like to see some references indicated that these combinations are even possible with the standards you suggest.

It's a little much to take seriously - no where near enough detail as to whether these components exist, or how much they'd cost.

-Steve

Jack Zhang
May 19th, 2005, 07:53 PM
I'd have expected MPEG-4 to be predominantly 4:2:2 or 4:2:0... but I don't know how well it scales up. Furthermore, why 50 Mbps?

Thanks, Steve, so it has to be 4:2:2 and 25 mbps to be an acceptable standard, that changes this: 60mins on DVCAM/DV only.

no where near enough detail as to whether these components exist, or how much they'd cost

IBM could create the codec processor and Sigma could create the "3" megapixel widescreen 3CCD system. You're right again, I've reconsidered for a 3 megapixel widescreen 3CCD system.

Thanks and double thanks for revising my thoughts!

P.S This will not be a big and bulky pro cam. It will be a Z1 look alike.

Lawrence Bansbach
June 15th, 2005, 03:17 PM
Well, I was wondering:

Is the combination of 48-bit colour and 4:4:4 even POSSIBLE in MPEG-4?
From the Wikipedia entry "H.264/MPEG-4 AVC (http://en.wikipedia.org/wiki/MPEG-4_Part_10)": "The JVT [Joint Video Team] recently completed the development of some extensions to the original standard that are known as the Fidelity Range Extensions (FRExt). These extensions support higher-fidelity video coding by supporting increased sample accuracy (including 10-bit and 12-bit coding) and higher-resolution color information (including sampling structures known as YUV 4:2:2 and YUV 4:4:4). Several other features are also included in the Fidelity Range Extensions project (such as adaptive switching between 4×4 and 8×8 integer transforms, encoder-specified perceptual-based quantization weighting matrices, efficient inter-picture lossless coding, support of additional color spaces, and a residual color transform). The design work on the Fidelity Range Extensions was completed in July of 2004, and the drafting was finished in September of 2004."

Jack Zhang
June 16th, 2005, 05:23 PM
I didn't know that, thanks! Now it will be 36bits of color with 4:4:4 sampling loosly compressed into H.264 (MPEG-4 Part 10).

Matthew Wauhkonen
July 7th, 2005, 10:00 PM
It cooks lunch for me, has a fifty gajillion bit cinegamma gigapixel sensor with 512 bit dynamic range, and costs about fifty dollars.

While this is not my first post, I'll label it my "most wildest" as well.

Jack Zhang
July 10th, 2005, 05:05 PM
Mine's way more practical, your's isin't. Portable HQ HD @ 1080p60 is definitly the future! look below at my signature!

Jack Zhang
August 2nd, 2005, 09:15 PM
* Not 3CCDs but a 3 Next-generation CMOS system for maximum dynamic range.

Noah Yuan-Vogel
August 3rd, 2005, 01:31 PM
encoding to a custom h.264 16bit 4:4:4 1080p60 stream realtime with portable hardware.... how? And then editing with it? On what?

Matthew Wauhkonen
August 4th, 2005, 09:54 AM
I'm pretty sure this post is a joke, parodying people's wild imaginations regarding yet to be released products with impossible specs.

At least I hope and pray that it is.

Jack Zhang
August 8th, 2005, 03:06 PM
Matt: Nope, I'm just trying to get Sony's attention, this isin't a joke! (after some fine tuning it won't be)

Noah: The next-gen Final Cut Pro might be a suitable editing platform.

Noah Yuan-Vogel
August 8th, 2005, 06:25 PM
so basically an updated viper hooked up to a made-up vtr that uses unreliable media and heavy compression? still a good $50,000+ over my price range if it existed, hard to get excited about. next gen fcp doesnt exist but when it does, I suppose it would be just HDVs style to use an (then) out-of-date compression scheme that does a poor job for acquisition just so consumers can still use their decade-old minidv tapes.

So you are saying in the future most people will still be digitizing from tapes in real time? What IBM 64-bit processor would it use?

Jon Kamps
August 9th, 2005, 09:14 PM
why not just drop the tape all together and go with a module that contained a pair of 120gb Laptopdrives (in raid 1 for redundancy). get rid of all this stupid mpeg crap (temporal compression is bad mmmkay) basically caputre HDCAM to the HD. I really think 4:4:4 is overkill right now. (hell look at starwars video quality wise that looked pretty good and from what i'm told it was all shot 1080p HDCAM which is 4:2:2.)

Noah Yuan-Vogel
August 9th, 2005, 10:16 PM
seems to me it might be a while before they put 3 8mp ccds in a prosumer camera. Why not just call this a wishlist. A lot of the specs don't seem like they make sense in the same camera. Most people who are concerned about datarates and harddrive space enough to use heavily compressed mpeg4 probably would prefer to spend harddrive on higher resolution or less compression than on 48bit color depth and 4:4:4 sampling.

Jack Zhang
August 13th, 2005, 02:43 PM
You're all right. This impossible idea might not be possible until 2010, so let's try to perdict what they might come up with. If they do think of this, some of the specs would be toned down.

Jack Zhang
September 24th, 2005, 07:23 PM
With the XLH1, we're still a baby step to true 1080p60. Sony's next HD cam must not dissapoint! They know that they've got more competition and they must work hard to get some of their profits back.