or Connect
AVS › AVS Forum › Video Components › Video Processors ›  DVDO iScan VP30
New Posts  All Forums:Forum Nav:

DVDO iScan VP30 - Page 8

post #211 of 6818
HD deinterlacing??

1080P24sF support?
post #212 of 6818
Quote:
Originally Posted by thebland View Post

HD deinterlacing??

1080P24sF support?

I assume that if it has hd de-interlacing, then they wouldnt be shy about it, and it would have a top position on their marketing.

Per Johnny
post #213 of 6818
Yeah. Josh's participation here is enormously appreciated, but I think we can learn as much from what has *not* been said as from what has. Don't get me wrong - I have an HD and an HD+ (in two separate theaters) and I'm a pretty happy DVDO owner. Love the service, love the company.

However, I really want my next scaler to be able to handle my next display...

TM
post #214 of 6818
Thread Starter 
The SiI504 is being used for deinterlacing so the VP30 will not support true 1080i deinterlacing. As has been stated by others, like Ofer, we are working on our next-generation deinterlacer which we should have displayed at our booth at CEDIA.
post #215 of 6818
Thread Starter 
Quote:
Originally Posted by thebland View Post

1080P24sF support?

Not at launch.
post #216 of 6818
Thread Starter 
Quote:
Originally Posted by DanHouck View Post


Josh, would there be any DVI-A output through the HDMI port of the VP30?

For an analog output, I would use the BNCs and then use a BNC-to-DVI-I dongle, if your projector does not have a standard analog input.

There is no analog signal on the HDMI output.
post #217 of 6818
So, Josh, can you tell us more about the next de-interlacers time frame? Sorry, thought I'd give it a shot. I'm hoping that your upgrade policy continues to be just as good as it was this time. Figure I'd get more value jumping on the VP30 now and then upgrading when the next gen comes around.
post #218 of 6818
Thread Starter 
We absolutely plan on continuing our upgrade program so that our customers have access to the latest and greatest ABT technologies. We do not have a specified time frame for the implementation of this next-generation deinterlacer (and honestly, if we did I wouldn't be able to tell you).
post #219 of 6818
Thread Starter 
I do have one correction to make to the datasheet. It specifies 12-bit DACs, they are actually 10-bit, 300 MHz instrumentation-quality DACs, with up to 10x oversampling and 2X oversampling for 1080p. I do not believe that any other video processor 2x oversamples 1080p.
post #220 of 6818
Hai Josh,

Just to be sure, is the sofware for the vp30 based on the work done on the 2.91 release (so is this the 3.0 release). In other words can we expect the new features added in 2.91 to be available on the vp30 (interlace over hdmi and vertical image shift come to mind).

And am i correct to suspect this is a merge of the old HD+ and denon 5910 code (that was also 10 bits). Last question is there enough extra room on the fpga's to expect more features in the vp30 platform (does it have room to grow).

Hope you can reply on a few of these keep up the good work,

Daniel.
post #221 of 6818
Quote:
Originally Posted by AndreYew View Post

But they throw away the four lowest values for every number. How do you distinguish between 1023 and 1022? You can't. In fact, you cannot distinguish between 1023, 1022, 1021, and 1020 since the top 8 bits will all be the same. In a fixed-point representation, resolution is exactly the same as dynamic range.



OK, think about this another way: 8 bits can unambiguously represent 256 discrete levels. 256 is much less than 1020. So how is 1020 levels 8 bits?

--Andre

Understood. What I was getting at was the lowest two bits contain
a fraction of 1% of the total 10 bit quantity measured and that dropping
them amounts to a fraction of 1% error. It was very late last night when
I was posting and I was not thinking clearly.
post #222 of 6818
Quote:
Originally Posted by darinp2 View Post

Put simply, how are you claiming that 5 and 6 are encoded in this case?

--Darin

I see the error of my ways :-) It was very late and I was tired when I
wrote that.

What I was getting at is the dropping the two LSBs of a 10 bit number
for processing amounts to less than 1% total error.

I claim temporarily insanity for trying to map 1020 distinct
numbers into 256 distinct states :-).
post #223 of 6818
Josh,

If you're allowed to answer this, will the next-gen deinterlacer make us video 480i fans happier, or should we not expect better performance than a Faroudja 23xx or Sil504 kind of solution?

--Andre
post #224 of 6818
Thread Starter 
AndreYew - I think it is safe to say that our next generation deinterlacer will be a giant-killer (but what else would you expect from ABT)
post #225 of 6818
Quote:
Originally Posted by tonydeluce View Post

I see the error of my ways :-) It was very late and I was tired when I
wrote that.

No problem. I'm sure I've done worse. I just figured that trying to code 5 and 6 would make you see it.

--Darin
post #226 of 6818
Quote:
Originally Posted by Josh@dvdo View Post

AndreYew - I think it is safe to say that our next generation deinterlacer will be a giant-killer (but what else would you expect from ABT)

Give Dale and the team my best wishes on getting it completed. We need more good deinterlacing in the world.
post #227 of 6818
Josh,

Thanks! Now I'm really looking forward to seeing what you have at CEDIA.

--Andre
post #228 of 6818
Quote:
Originally Posted by steviec View Post

So can we safely say the new VP30 is same product as the HD+ but ,different case with the addition of HDMI?

Not quite.



Quote:
Originally Posted by Josh@dvdo View Post

Here are the main differences between the VP30 and the HD/HD+:

4 HDMI Inputs and 1 HDMI Output (Audio/Video)

HD and HD+ have 1 DVI input and 1 DVI output

Precision Video Scaling II - 10-bit scaling, non-linear scaling, enhanced sharpness control

HD and HD+ use Precision Video Scaling (I) which is 9-bit

12-bit Video Decoder (480p/576p/720p/1080i)

HD and HD+ use 10-bit Video Decoder (VP30 does for 480i and 576i)

BNC input and output capable of RGBHV or Component

HD and HD+ have HD15 connectors capable of RGBHV or Component

RGBHV Input is processed and transcoded

HD and HD+ pass through on HD15 input

Component HD is processed and transcoded

HD and HD+ pass through 720p and 1080i on Component inputs

Elegant front panel design
Analog Audio Input
post #229 of 6818
Quote:
Originally Posted by steviec View Post

Originally Posted by Josh@dvdo
AndreYew - I think it is safe to say that our next generation deinterlacer will be a giant-killer (but what else would you expect from ABT)

I think it is alittle late and that this is what the VP30 should have been.
Just my opinion.

A little late compared to whom, and what?
post #230 of 6818
Quote:
Originally Posted by steviec View Post

Here Stevie.

Sorry, not impressed since the Sil 504 is still used.

What would you have preferred? What's better for 2 grand or less?
post #231 of 6818
I think, compare to the current HD+, the upcoming VP30 has the following differences:

- new (better?) 12bits A/D for analog video input
- new (better?) scaling chip
- 1080i/720p process over component/RGB input
- RGB/component transcoding
- HDMI input/switcher
- analog audio input
- BNC (better?) connecter for RGB/component output

Deinterlace (both SD/HD) performance should be the same as HD+ with the same Sil504 chip.

A minor question: can the BNCs component/RGB output and the HDMI output be used at the same time?

regards,

Li On
post #232 of 6818
Quote:
Originally Posted by steviec View Post

Sorry, not impressed since the Sil 504 is still used.

I feel the same way.

At least we can be confident that there'll be a VP40 coming along within a year or so...

Quote:
Originally Posted by Q of BanditZ View Post

What would you have preferred?

Something with proper diagonal interpolation for video material and motion adaptive + inverse telecine deinterlacing for 1080i.
post #233 of 6818
Quote:
Originally Posted by Carled View Post

Something with proper diagonal interpolation for video material and motion adaptive + inverse telecine deinterlacing for 1080i.


Can all that be done for 2 grand or less? By anyone?
post #234 of 6818
Quote:
Originally Posted by Q of BanditZ View Post

Can all that be done for 2 grand or less? By anyone?

The KDS Leeza already does, excluding the diagonal interpolation, and that's been round a while.
post #235 of 6818
Quote:
Originally Posted by Josh@dvdo View Post


There is no analog signal on the HDMI output.

Are you saying that any analog inputs to the VP30 would not be scaled/passed to the HDMI output?


Thanks
post #236 of 6818
Quote:
Originally Posted by vfrjim View Post

Are you saying that any analog inputs to the VP30 would not be scaled/passed to the HDMI output?

"There is no analog signal on the HDMI output" doesn't imply that, at least to my understanding. His quote doesn't rule out the VP30 digitizing the analog input, and scaling/passing it out digitally through the HDMI output.
post #237 of 6818
Quote:
Originally Posted by vfrjim View Post

Are you saying that any analog inputs to the VP30 would not be scaled/passed to the HDMI output?

No, he's saying that the HDMI jack won't opperate like DVI-I.

Analogue->Digital transcoding has already been confirmed.
post #238 of 6818
Getting rid of jaggies is the first stage, and that's what diagonal interpolation does, but I'd like to see the loss of resolution with moving objects addressed, too, in an ideal world. Motion-compensated deinterlacing would be nice, but probably not practically doable for a typical consumer budget in the next 5 years (I'd be happy to be wrong).

BTW, does anyone know: is Dale Adams one of the original Sil504 designers? I imagine they must have learned quite a bit since that chip came out, and it will be interesting to see what they do next.

--Andre
post #239 of 6818
Quote:
Originally Posted by AndreYew View Post

Getting rid of jaggies is the first stage, and that's what diagonal interpolation does, but I'd like to see the loss of resolution with moving objects addressed, too, in an ideal world. Motion-compensated deinterlacing would be nice, but probably not practically doable for a typical consumer budget in the next 5 years (I'd be happy to be wrong).

MC deinterlacing requires super-computer processing power. Hopefully Cell processors might make it possible in the future, but at the moment it can't be done in consumer grade FPGAs.

Quote:


BTW, does anyone know: is Dale Adams one of the original Sil504 designers? I imagine they must have learned quite a bit since that chip came out, and it will be interesting to see what they do next.

That's right.

I certainly can't wait.
post #240 of 6818
The Teranex people say they do a full-exhaustive search of every pixel to do their motion compensation: I assume this means they compare every pixel with every pixel in the next field to determine the motion vectors, which seems like it should yield good results, albeit at a very high cost. Figuring that the big Teranexen cost about $80K, and the cost of computing power comes down 10x every 5 years, I came up with my WAG of 5 years before we get a consumer box that can do something comparable.

More WAGing: if Josh says they're killing giants, and giants are companies like Teranex, then maybe we'll get something cool sooner.

Thanks for the info on Dale.

(WAG = Wild-A**ed Guess)

--Andre
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Video Processors
AVS › AVS Forum › Video Components › Video Processors ›  DVDO iScan VP30