[rogo]

Ah, September 2004... We are intoxicated by the possibilities of Realta and Gennum's VXP.


A year has passed quite nearly.


Not only are both essentially no shows (yes, I know about the Denon), but...


... the HDMI DVDO variant isn't around (is it even announced?) ...


... nothing particularly exciting has shipped from anyone else (Key, Lumagen, etc.) ...


... and to make it worse, I still can't buy a reasonably priced multi-in, multi-out transcoder that supports 4 HDMIs and various analog sources, all with audio ...


In fact, I can't even buy an unreasonably priced version of that. It doesn't exist.


I have to believe I'm not the only person with money to spend in this segment and no products to buy.

 

[tonydeluce]

Quote:

Originally Posted by rogo Ah, September 2004... We are intoxicated by the possibilities of Realta and Gennum's VXP. A year has passed quite nearly. Not only are both essentially no shows (yes, I know about the Denon), but... ... the HDMI DVDO variant isn't around (is it even announced?) ... ... nothing particularly exciting has shipped from anyone else (Key, Lumagen, etc.) ... ... and to make it worse, I still can't buy a reasonably priced multi-in, multi-out transcoder that supports 4 HDMIs and various analog sources, all with audio ... In fact, I can't even buy an unreasonably priced version of that. It doesn't exist. I have to believe I'm not the only person with money to spend in this segment and no products to buy.

I am waiting on the new DVDO - I understand it is right around the corner and will

be announced shortly...

 

[mark haflich]

At least several products of interest here will be announced by existing processor manufacturers. However, an announcement and a dummy box does not make a product. I doubt anything announced to be announced would become available for a long long time.


For example, it has long been speculated that Lumagen will announce say a new Realta based processor, not as a replacement to its existing processors to which features are still being added and which I expect continue to be added, but as a much more expensive higher end processor. Given the length of time needed for Lumagen to develop the software for its current generation to provide promised features, one would expect considerable developmental time for the programming necessary for any new processor.

Just the licensing time necessary to get a product with HDMI is long, aka part of Algolith's delay. And remember the Dragonfly is essentially a board design provided by Silicon Optix in a box and it still isn't here because of software developmental issues. Fixing glitches is easy. Developing solutions to non glitch problems is quite another.

 

[oferlaor]

rogo,


The more complex the designs - the longer it takes for products to come out. If you remember, there was a similar period of waiting a few years ago as the original Rock was coming out and we were very hopeful about the possibility of an open system that could compete with that current crop of scalers.


Based on what I'm hearing we should see and hear very interesting things in Sept (CEDIA). It's too bad I can't attend (yet again) - this time I have a really good excuse, my wife is due with our third child around that time....

 

[AlanMFriedman]

Quote:

Originally Posted by oferlaor this time I have a really good excuse, my wife is due with our third child around that time....

That's the best possible excuse for being a "no-show". Congratulations.

 

[mark haflich]

To steal one of Ofer's lines, Hey that is really great news. Give her the best from all of us.

 

[mburnstein]

yep

 

[oferlaor]

will do

 

[rogo]

Quote:

Originally Posted by oferlaor rogo, The more complex the designs - the longer it takes for products to come out. If you remember, there was a similar period of waiting a few years ago as the original Rock was coming out and we were very hopeful about the possibility of an open system that could compete with that current crop of scalers. Based on what I'm hearing we should see and hear very interesting things in Sept (CEDIA). It's too bad I can't attend (yet again) - this time I have a really good excuse, my wife is due with our third child around that time....

I guess, Ofer, I'm more persuaded by Mark's comments than making excuses for these people.


If a Silicon Optix-development board can't be made into a product in less than a year how can I get excited about anything I hear at CEDIA?


Stuff we can buy is what interests me. And, yes, I have a feature set I'm demanding. And it's driving me crazy no one is offering such a thing. Nor do I expect to see one anytime soon... Alas.

 

[Dale Adams]

Quote:

Originally Posted by rogo Stuff we can buy is what interests me. And, yes, I have a feature set I'm demanding. And it's driving me crazy no one is offering such a thing. Nor do I expect to see one anytime soon... Alas.

What would that feature set be?


- Dale Adams

 

[Kei Clark]

He wants multiple analog/digital video scaling/switching and audio switching in one box.

 

[mark haflich]

Most of these video processor companies are very small and have VERY VERY limited resources. Take Taw back in the old days. What did it have? One programmer, a part time sales manager who essentially worked for free product, an owner and his stepson and a couple of volunteer helpers at shows?


Take Lumagen. I think a much more successful company How many programmers does it have? Two? Three? I do no know but very few. If you have massive resources you have a better chance of faster developmental times as well as shorter production times.


Algolithis initial production run of Dragonfly's is 100. Using an existing board design with some very very needed improvements. Why the delay? One would expect algorithm developmental solutions and then code writing. This takes time and resources and remember the solution must be done in a manner that the machine can handle. Speed etc. Hey Boss, I have a solution but the machine would need 2 minutes to crunch the program and we have 2.1 nanoseconds.

 

[DanHouck]

Quote:

He wants multiple analog/digital video scaling/switching and audio switching in one box.

Me too!

 

[Dean Roddey]

Quote:

Hey Boss, I have a solution but the machine would need 2 minutes to crunch the program and we have 2.1 nanoseconds.

You're promoted! Ship that puppy!!!

 

[flyingvee]

Quote:

Originally Posted by mark haflich Speed etc. Hey Boss, I have a solution but the machine would need 2 minutes to crunch the program and we have 2.1 nanoseconds.

now THERE is a lip-synch problem. you can optimize code, you can throw faster processors at the problem, but some things still take time (2 minutes seems a bit long, but we get the idea.) Course, if they could put a 2 minute audio delay into the box also, then the only problem would be waiting 2 minutes for the machine to "boot."

 

[RichB]

There are some advantage to the smaller company building scalers. Companies like Lumagen and DVDO are responsive to customers.


If you were to buy scaler from NEC, would there be any one answering questions on this forum?. Would they care if it does not sync properly with Panasonic or Sony display or projector.


-- Rich

 

[rogo]

Quote:

Originally Posted by Dale Adams What would that feature set be? - Dale Adams

Dale, I don't mean this quickie one off to be an exhaustive list, but I'll start and let others chime in.


Design goal: Enough inputs and outputs to drive the entire audio and video chain. The box is not an amplifier and it's not a display. But it needs to switch all the audio and video, it needs to synch any audio with the video and it's from a video processor company so it should process video.


Video processing features: Scales any reasonable known resolution to at minimum a fixed set of output resolutions including 1280 x 720, 1366 x 768, 1920 x 1080. I suppose standard 480p output and standard 1080i output are musts just in case.


Deinterlaces using whatever algorithms you'd like but has to be able to handle 3:2 pulldown of 1080i to make 1080p. Motion adapative or compensated stuff for 1080i is also nice. These scaling and processing features merely have to be done well or not at all. If it did the rest, you could drop those and leave them to the display and just pass through the video per the requirements below....


Transcodes all video to HDMI. Composite, S-video, component get sent out the HDMI(s). This is really the single most essential feature. The goal here is to organize the sources and have a single switch point.


Inputs:


* 4 digital inputs. At least 3 are pure HDMI. The fourth can be DVI or HDMI.

* 2-3 component inputs. If people have that many, this box has to be the "one box that switches them all". These should probably be VGA compatible and therefore offer 5 inputs.

* 2-4 composite / s-video inputs

* Every single input has sound. So every single input needs its own Toslink or digital coax to correspond with it. The HDMI must pass audio, but since some sources are DVI, this means they also need a corresponding audio. To make things simpler on conncetors, it's fine if you have "mapped" audio and offer 8-12 total connectors, maybe 9 optical, 3 coax. I don't care a whit about coax, but I know some people do.


Outputs:


*2-3 digital outputs and really, matrix capability would be incredibly nice. Distributing HD video is pretty impossible today so the 2nd or 3rd outpu could feed either your 2nd display (think plasma and projector) or something else like a 1 x 4 distribution device for HDMI to send at least one other HD picture to at least one other TV in the house. Build the distribution amps as companion pieces that have some extensible remote control architecture and I'll pay for those too.

* Equivalent audio outputs for the same reason


That's my "straw man". I'm willing to refine it with the help of others.


Oh, right, price, huh? I'd like it for $1499, which I know is not going to happen. How about we discuss price in the context of what's realistic. And keep in mind, you can take out all the video processing save the transcoding of the analog. Yes, you can remove it. Gonzo. Sell it to me separately if you wish. Sell two versions of the box, one with, one without. I don't care. My TV can deinterlace and scale. OK, it doesn't do it as well as DVDOs stuff. Not close. But it can do it. Sell me the transcoder / all-input switcher for $999 and offer me a companion scaler for $1999 (or whatever the prices need to be to make your typical margins which I'm not complaining about).


Mark

 

[Carled]

I certainly agree with the call for more inputs on video processors. The designers seem to be intending them for idealistic situations, but in the real world it's not uncommon to have a whole bunch of devices, especially when it comes to game consoles, multiple VHS decks, mutiple set-top-boxes, and multiple compters. And the last thing you want when you fork over the big $$$$s for a scaler is to also have to worry about switchboxes with sufficiant bandwidth, finding one that is remote opperated, and then loading the controls into a learning remote or RS-232 feed. I just want to push a single button on my remote and be go.


I'd like to see ~6x HDMI, ~4x component, ~2x SDI, (prehaps an additional ~2x HD-SDI if it's actually used), ~3x RGB, ~3x S-video, ~3x composite, ~2x SCART, and ideally an IEEE-1394 input for camcorders, etc.


There seems to be this phycologicl barrier against making scalers bigger than one rackmount space in height. I wouldn't mind something 4RU (or even bigger), if it did what I need of it.




On other thing while I'm at it. Scalers should support scaling of computer resolutions, in my opinion. It's nice to be able to feed your computer through your scaler to be able to implement image adjustments, pillarboxing, image shifting, and nailing 1:1 mapping.

 

[Dean Roddey]

Just for the sake of argument, if you were starting to design something now (already late in the game) and delivering it maybe next year, is composite really of interest anymore in a device like this that's like to end up in a fairly high end theater? I'd almost say the same about S-Video. Is anyone likely to buy such a device going to care about those? I'm not even high end and I've not had an S-Video or composite based device in my home theater for years.


Maybe composite/S-Video are such that they are just a gimme because they are supported by the chips used anyway. But if it reduced the cost by 10%, for me I'd say toss'em. Maybe it'd cost you some sales, but if it made it smaller, lighter, cheaper, and less complex, and it could serve 80% of the market for it (mid-high end home theater I'd think?), I'd argue for small, lighter, cheaper, and less complex.

 

[Stacy Huff]

I'd like to be able to use any new scaler with my LD player, so I'd still need composite and S-video.