Recording HD from HDD200 Data Port? - AVS Forum
Forum Jump: 
 
Thread Tools
Old 12-21-2000, 01:09 PM - Thread Starter
CSM
Member
 
CSM's Avatar
 
Join Date: Aug 2000
Location: Vancouver
Posts: 100
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Would it not be possible to easily record the HD "signal" sent to the General Instruement's HDD200 data port for transmission to it later?

Save the data stream to a hard drive or tape?

No need to "decode" the data--just want to store it and then transmit it to the decoder at some other time.

Any ideas or am I missing some obvious obstacle / technical issue?

CSM is offline  
Sponsored Links
Advertisement
 
Old 12-21-2000, 01:18 PM
AVS Special Member
 
David Panko's Avatar
 
Join Date: Nov 2000
Location: Toronto, Canada
Posts: 1,310
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Should be possible.. there was a short thread discussing this in the HDTV recorders forum.
http://www.avsforum.com/ubb/Forum20/HTML/000131.html

We need an EE to take a crack at it.

Dave.
David Panko is offline  
Old 12-21-2000, 02:05 PM
AVS Special Member
 
JohnnyG's Avatar
 
Join Date: Dec 1999
Location: Toronto, Ontario, Canada
Posts: 1,581
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 10
You first have to figure out what format the data is being sent in. No easy task.

------------------
John Golitsis
Next Big Thing Electronics
Mississauga, Ontario, Canada
http://www.nbtelectronics.com

[This message has been edited by JohnnyG (edited 12-21-2000).]
JohnnyG is offline  
Old 12-21-2000, 03:06 PM
Newbie
 
mdavies's Avatar
 
Join Date: Jun 2001
Location: Mission, BC, Canada
Posts: 2
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally posted by JohnnyG:
You first have to figure out what format the data is being sent in. No easy task.

I was the one who originally posted the thread in the HDTV recoding group.

After doing a little research and looking at the signal with an oscilliscope, here was what I found:

There are 2 signals and one ground. The two signals are differential (signal1 - signal2 = 2*signal1), so they contain the same information.
The signal is a 29.4 Mbps Alternate Mark Iversion (AMI) encoded serial data stream. AMI is a serial format where bit values of 1 alternate between a positive voltage and a negative voltage, with bit values of 0 having 0 voltage. So a bit pattern of 1001011001 would look like this:

+00-0+-00+

+ = positive voltage
- = negative voltage
0 = 0 voltage

Even though I am an EE, I primarily only write software, so the hardware portion of translating this signal into firewire or something else a PC can recognize is beyond my abilities. If someone else can figure this out, I would be happy to contribute to writing the software.
mdavies is offline  
Old 12-21-2000, 10:03 PM - Thread Starter
CSM
Member
 
CSM's Avatar
 
Join Date: Aug 2000
Location: Vancouver
Posts: 100
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Thanks for the responses..Yep--sounds like we need a hardware guy first..Don't imagine that GI would like to make/sell a nice add-on for timeshifting?
CSM is offline  
Old 12-22-2000, 09:45 AM
AVS Special Member
 
Man E's Avatar
 
Join Date: Oct 1999
Location: The Wilderness Near Hell
Posts: 4,460
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Have you tried Tom or one of the other hardware guys at the 4DTV forum? Last I heard over there, it wasn't known to be possible, but you might have some new information that could assist in the definition of such a creature. I'm in if it's doable.


------------------
Please visit www.rettsyndrome.org

HBO is guilty of Crimes Against Filmanity!

From Paragraph 44 of the 5th Report and Order: We note in this regard that broadcasters and networks have emphasized their commitment to high definition television.

Our Silent Angels Please visit The Manny Page!
Man E is offline  
Old 12-22-2000, 11:08 AM
AVS Special Member
 
JohnnyG's Avatar
 
Join Date: Dec 1999
Location: Toronto, Ontario, Canada
Posts: 1,581
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 10
Great work mdavies! You mention converting the data to Firewire, but how about IP? There's a lot more expertise out there in that area.

What are your thoughts on that Rick? Would your idea be 'easier'?

------------------
John Golitsis
Next Big Thing Electronics
Mississauga, Ontario, Canada
http://www.nbtelectronics.com
JohnnyG is offline  
Old 12-22-2000, 12:11 PM
 
wirehead_rick's Avatar
 
Join Date: Jul 2000
Location: Streamwood, IL
Posts: 444
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally posted by JohnnyG:
Great work mdavies! You mention converting the data to Firewire, but how about IP? There's a lot more expertise out there in that area.

What are your thoughts on that Rick? Would your idea be 'easier'?
What is IP? Could you possibly mean "Internet Protocol"? If it does, it doesn't really address the issues at hand. From what I know, IP as you say doesn't even relate to hardware other than it usually takes hardware to implement it. As such the internet protocol can be implemented on all types of hardware (PPP, ethernet, X.25, ISDN, RS-232, ATM, etc.) but does not address the specific issue. That is converting a "generic" if you will, 29.4Mbs serial bit stream to a binary file image on a hard drive and being able to play it back in the same order it was recorded.

Apologies if IP means something else that doesn't strike me at the moment.

Rick
wirehead_rick is offline  
Old 12-22-2000, 12:39 PM
 
wirehead_rick's Avatar
 
Join Date: Jul 2000
Location: Streamwood, IL
Posts: 444
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
New idea.

Maybe the 29.4Mbs bitstream could be AM encoded and converted into a video signal with off the self chipsets? Any other EE's want to chime in on this idea?

I remember about 15 years ago Samsung (maybe Hitachi too?) came out with a digital audio converter that would take a stereo analog signal and convert it to digital. It would then modulate the resulting digital bitstream onto a baseband video signal and output the video signal to be recorded on to a standard VCR. You could then play back the video signal back into the Samsung converter and listen to your recorded audio. This was all before HiFiVHS or HIFiBeta and before DAT, so it was the only alternative for audiofiles to record their own digital audio.

Available today are multitrack digital audio recoders designed for use in home studios which will record and play back up to 8 audio tracks onto a SuperVHS or HI8 tape.

A stereo digital audio bitstream has a bitrate of:

44.1Khz x 16bits/sample x 2(stereo) = 1.4112Mbs.

and 8 Mono digital audio streams has a rate of ~5.6Mbs.

A lot less than 29.4Mbs but with an AM encoder/decoder it might be doable. Say you use a 5 bit DAC which gives you a ratio of 32/1 and a dynamic range of ~30db (I think VHS can handle this and most probably S-VHS can). The 32/1 reduces the 29.4Mbs under the ~1.4Mbs rate, so it now falls in the realm of real possibility since it has been already done before.

If off the shelf video chipsets were used, it might be possible but it is definitely not an area of my expertise (analog video). I can do digital, but I see clock recovery and syncing off the played back VHS signal being the scariest part of the design.

Overall what does all this mean? A black box between the GI and your standard VCR to record and playback the bitstream signal. Would allow the VCR to function as it normally would (time-shifting/cheap archiving/trading of tapes [did I say that?]/etc.).

Like most ideas off the top of my head it is a lot more pie-in-the-sky than my other idea (PCI CARD) but might be just as doable.

Both implementations would serve different feature sets, obviously.

Rick

wirehead_rick is offline  
Old 12-22-2000, 10:16 PM
Senior Member
 
davidahn's Avatar
 
Join Date: Oct 2000
Location: Apple Valley, California, USA
Posts: 264
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 16
I am NOT an engineer, but that was exactly what I was imagining as a possiblity (modulation onto analog video tape); what say someone finish this idea off and make himself some serious bucks?

Consider encoding into formats already present:

29.4 Mbps = 3.675 MB/s

For example: mini-DV is digital and has throughput of 5.9 MB/s, and all you'd need is hardware to encode/decode the firewire in/out from any mini-DV camcorder or VCR.

No need to reinvent the wheel here!

David
davidahn is offline  
Old 12-22-2000, 10:18 PM
Senior Member
 
davidahn's Avatar
 
Join Date: Oct 2000
Location: Apple Valley, California, USA
Posts: 264
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 16
One other thought:

Since satellite receivers don't have Firewire out for recording devices, the only output you can use that ALL HD-capable receivers have is the VGA/DB-15 output. It seems to me that this can be modulated also, but would require much more work. Anyone given any thought to this?

David
davidahn is offline  
Old 12-22-2000, 10:42 PM
 
wirehead_rick's Avatar
 
Join Date: Jul 2000
Location: Streamwood, IL
Posts: 444
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
I am a hardware guy with some (not a lot) of PC hardware experience but do have a good deal of experience in other areas. I have designed an Ethernet transciever FPGA/IC and have experience with serial data processing and the PCI bus as well.

This is not a slam dunk as far as I can tell. I can't think of any quick and easy ways to get a 29.4Mbs data stream into a PC. Maybe using firewire but I know that firewire has a fairly complex protocol system in place to handle it's "Plug and Play" features, so what you save in hardware there gets replaced by complexity of design. A "simple" interface would probably need some sort of processer at a minimum to handle the protocol. There may be specialized chips out there to handle such a task (probably are), but I'm not sure and have no direct knowlege of them. I can research this.

If I was asked to design such a system I would probably use an FPGA on a PCI card and just customize the serial interface and use a standard PCI bus protocol transfer the serial data into the system in PCI-Bus size chunks, or maybe using a DMA like setup to go straight to memory. Software would have to take control of the PCI to system operation (as in "driver"). The FPGA vendor I would choose provides standard PCI bus protocol IP blocks that you can drop into a design, so that wouldn't have to be designed from scratch.

The serial data format described (AMI) would certainly make designing a clock recovery circuit (absolutely necessary) easy enough to do. And serial to parallel conversion along with buffering is not too difficult in the world of FPGA's. If the GI data stream is not necessarily blocked (frame based), there's no reason it shouldn't work (it might work anyway, even if it is frame based).

Still, doing a H/W project like this is no easy task. I would charge quite a bit of money to do such a system if I was still doing consulting work.

I am currently doing IC design so my FPGA and PCB design skills are a bit rusty (2 to 3 years out since last design, which is ancient in this industry) and I don't have access to any tools (except PCB design software). BUT I might consider looking into the feasability of taking on a project like this (it actually sounds like fun) if I had some help from the members here.

First, I would need a General Instrument HDD200, along with any other HW for testing and evaluation. I have no such animal and don't know where to get it. I DO have an HDTV and a DTC-100, so I'm not completely without HD capability. I would need it after delivery of the first PCB's.

Second, I would need FPGA's and the development tool kit. I have friends in the industry who might lend me some tools for a short while and free-sample me some FPGA's. It might be contingent on me not making any money off the project in the future (I must admit, the mind does wander in this area!) and buying future FPGA's.

Third, I would be able to supplement the cost of first PCB's and second PCB's but after that member participation financially in getting more systems built would be necessary.

Fourth, I AM really concerned about interfacing with proprietary protocols, licensing and copy issues. If GI gets wind of such a project, how will they react? If the MPAA gets wind of such a project, how will THEY react? I really would hate to complete such a project and then find myself being hauled off to court with nobody to stand next to or behind me (or help with any legal costs)! Any ideas in this area would be welcome.

Just one quick question to mdavies, can you do the driver along with any application software, or can you only do the application? I could probably do a driver, but my skills in this area are pretty rusty. And I have never written a driver for Windows. Also, I think you need to access to many proprietary M$ development tools and API's to do a decent Windows driver. None of which I have.

Well, there it is. All in all I'm interested, but I have experienced many a bright idea in the past only to see it come crashing down into the jaws of reality. People in general don't appreciate the amount of work and effort goes into designing something say like an IDE interface. We users only reap the benefits when we just go to Best Buy and slap in a hard drive or CDROM drive and it magiaclly works. That, my friends, is not an easy task to do, even if it seems easy and simple by observation.

This idea is definitely possible and is definitely do-able. I really see the big issues, even if I don't do it, being with the copyright stuff, and possible percieved encroachment on any protocol patents GI might have (even if there is no attempt to reverse engineer the protocol itself).

My $.02 worth.

Rick
wirehead_rick is offline  
Old 12-22-2000, 11:44 PM
Senior Member
 
davidahn's Avatar
 
Join Date: Oct 2000
Location: Apple Valley, California, USA
Posts: 264
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 16
I am NOT an engineer, but that was exactly what I was imagining as a possiblity (modulation onto analog video tape); what say someone finish this idea off and make himself some serious bucks?

Consider encoding into formats already present:

29.4 Mbps = 3.675 MB/s

For example: mini-DV is digital and has throughput of 5.9 MB/s, and all you'd need is hardware to encode/decode the firewire in/out from any mini-DV camcorder or VCR.

No need to reinvent the wheel here!

David
davidahn is offline  
Old 12-23-2000, 09:25 AM
AVS Special Member
 
imws's Avatar
 
Join Date: Sep 2000
Location: New York, NY USA
Posts: 1,399
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Just a quick aside here, 29 Mbs read 29 Megabits per second, I'm confused, why does Rick think this will be a problem to interface with a PC?

Getting a signal with this kind of bandwidth into a computer should not be problem since there are ethernet cards out there for the PCI bus that are rated at 1Gbs read 1 Gigabit per second. Bits are bits as they say.
imws is offline  
Old 12-23-2000, 01:10 PM
 
wirehead_rick's Avatar
 
Join Date: Jul 2000
Location: Streamwood, IL
Posts: 444
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally posted by imws:
Just a quick aside here, 29 Mbs read 29 Megabits per second, I'm confused, why does Rick think this will be a problem to interface with a PC?

Getting a signal with this kind of bandwidth into a computer should not be problem since there are ethernet cards out there for the PCI bus that are rated at 1Gbs read 1 Gigabit per second. Bits are bits as they say.
You do have a valid point.

Like I said, I have some but not a lot of PC experience. Although I have never seen 1Gb Ethernet. I have seen 100Mb Ethernet (EtherFAST). EtherFAST might be fast enough. Just remember that 100Mb is the flat rate. It doesn't include packet overhead that must be fit into same bandwidth. It does seem to me that there sould be plenty of room to fit 29mbs into a 100Mbs channel even with overhead.

A custom protocol would have to be devised, since Ethernet uses HDLC and MAC hardware interface protocols (FYI, if you are interested you can look up the standard ISO communications protocol stack model). This would not nearly be as difficult as Firewire would be. Still you need an intelligent controller to handle upper level protocols betwwen the 29Mbs stream and the ethernet stream.

The Ethernet controllers you plug into your PC already handle the lower level protocols so what ever you hook on the other side of the cable, has to talk the same language. You just can't shove bits down the cable and expect the controller to intelligently process it. The controllers are optimized to a set of standards that govern the data flow across the cable that has to handle a bunch of things (like collision detection and retransmision, etc. Ethernet IS a bus based protocol even if the electrical configuration is a star [hub based] based network).

All in all, it might be easier than an FPGA based system but my expertise makes me lean towards the FPGA on a PCI card.

davidahn:

Firewire is not a simple system even though it gives the appearance of being simple (An endearing attribute of firewire and I think points out some of it's greatest power). Yes the cable is only 6 or 4 wire, but the controllers at the end of the cable are very complex. Every time you unplug or plug in a firewire device into a Firewire network, everything on the entire bus resets itself, the bus ID's are completely re-configured, and the bus data traffic is re-started (even if it is only two devices and one cable). Theoretically you could be recording on your Firewire VCR a program from your Firewire tuner, while listening music from your Firewire digital radio reciever and if you plugged in your digital video camara, the reconfiguration would occur WITHOUT interruption of any of these other operations. That is the power of Firewire. And it is not an easy protocol to implement to achieve this. You could argue to implement a subset of the protocols, but both controllers on each end of the cable would have to know this. That would be difficult with an off the shelf firewire card you would plug into your computer.

As far as recording the RGB right off of the RGBHV output of a tuner, it is certainly possible. I'll do a quick calcualtion:

1080ix1920x60fps = 540x1920x60fps = 62,208,000bps

Not insane, but clearly more bandwidth is needed. A high quality ADC is needed (actually three), but simple compression can be implemented by taking advantage that color doesn't need the same bandwidth as luminance so I could see it being possible to get to 29Mbs pretty easily (Maybe that's what GI did?).

Of course, I have heard that it is exactly this kind of a recorder that makes the MPAA shake in their shoes. Seems to be a big reason why MPAA wants encryption through the entire HD signal chain so engineers like me can't get at the raw data (like we couldn't just hack into the projection system!).

Rick



[This message has been edited by wirehead_rick (edited 12-23-2000).]
wirehead_rick is offline  
Old 12-23-2000, 01:28 PM
AVS Special Member
 
Man E's Avatar
 
Join Date: Oct 1999
Location: The Wilderness Near Hell
Posts: 4,460
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally posted by wirehead_rick:
A black box between the GI and your standard VCR to record and playback the bitstream signal. Would allow the VCR to function as it normally would...
Is this similar to the, now defunct, Panasonic device that many are using to record from minidish systems? This would be fabulous since I already have an SVHS deck and those tapes are fairly inexpensive.



------------------
Please visit www.rettsyndrome.org

HBO is guilty of Crimes Against Filmanity!

From Paragraph 44 of the 5th Report and Order: We note in this regard that broadcasters and networks have emphasized their commitment to high definition television.

Our Silent Angels Please visit The Manny Page!
Man E is offline  
Old 01-28-2002, 06:23 PM
Member
 
spexman's Avatar
 
Join Date: Jan 2002
Location: Toronto, Ontario
Posts: 99
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Just to let you know that I would be VERY interested in purchasing the hardware/software combination that would allow us to record the digital bitstream onto a PC in such a way as to output it identically to the HDD 201 for decoding. Unfortunately, I have NO electronics or programming experience to help out. I see that a few of you do indeed and if you could somehow combine your efforts to get this to work, I would be delighted and also your first customer! Please keep me posted on any future developments as I am itching to be able to record HD! Thanks guys!
Spexman
spexman is offline  
Old 01-28-2002, 07:05 PM
AVS Addicted Member
 
Ken H's Avatar
 
Join Date: Nov 1999
Location: Metro Detroit
Posts: 45,876
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Moving to HDTV recording.

Please note that with the exception of the last post, this topic is over a year old. To my knowledge, nothing came of this idea.

'Better Living Through Modern, Expensive, Electronic Devices'

Ken H is offline  
 
Thread Tools


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off