HDMI output for A1200?

[alexh]

You could perhaps ask Jens schoenfeld to add a digital interface to his new Scandoubler with the ability to get the data before and after the scandoubler?

Dont forget that HSync and VSync are not available on the Lisa.

Jens has come up with a technique where he doesn't need them, but he's keeping the details close to his chest.

[alexh]

sigmason wrote:
Considering the Amiga must create all the signals it needs to produce the video output, I have no doubt that an analysis of the schematic after the Lisa chip should reveal all the logic that went into the production of the video signal.

Erm, lisa has RGB, alice has H&V. That is why classic scandoublers like DCE FlickerMagic clip over both chips. Like I said earlier, Jens Schoenfeld has worked out how to do without the Alice connections.

sigmason wrote:
Now consider that by taping the output side you are potentially creating yourself a new problem.

I dont think you have to wory about that. Lots of internal scandoublers do it without any problems.

sigmason wrote:
As to whether it's better to simply 'replace' the Lisa or deal with the output modes I'll leave up to the person that  tackles this.

You cannot 'replace' the Lisa.

[alexh]

Eriond wrote:
The really interesting question here is at what rate does this sampling occur?

~28MHz (same as a TV) for all Amiga modes except Euro72, Super72 and Multiscan. (Non of which you really need to support)

Eriond wrote:
However, the question is still; is it the graphic chip that dictates when each sample occur or is it the internal clock of the HDMI TX?

I would imagine the HDMI TX would slave to the gfx input.

If HDMI supports 576i then it should be fine.

[alexh]

Sniff... sniff... what's that smell?

You need to go away and read the HDMI spec again.

[alexh]

@sigmason

Wow nice post.

If we ignore Super72, Euro72, and multiscan the Amiga always drives an output compatible with a regular TV. That means the Lisa output is always compatible with these four modes : 720x240p, 720x480i, 720x288p & 720x576i.

And if we ignore interfield motion (I know we shouldnt) then the Lisa output is always compatible with 720x480i or 720x576i.

The actual number of changing pixels may be lower, but the DAC doesn't know that. It is just sampling pixels at 28MHz.

I would imagine we are just going to feed the RGB outputs + sync + clock + config to the HDMI transmitter.

As you say these modes are below the 25 MP/S and so (from reading the HDMI spec) it appears we may need to have the Pixel Repetition (PR0:PR3) fields set (which will just be a counter) but I wouldn't be surprised if the transmitter does this automatically when configured for TV video sources.

sigmason wrote:
Let's consider this: if the Amiga at NTSC HiRes interlaced (which is a fine mode and the Toaster and other genlocks should work at it) is spitting out 640x400x60Hz or 480i.

Erm sorta.

sigmason wrote:
The digital data you are tapping at Lisa is pixel data for less pixels then the HDMI interface will require in a given frame or field because the HDMI display modes are all more data then you actually have (causing you to do pixel repetition on HDMI).

Yes the Lisa pixel data is less than 25MP/s and (from what I read in the spec) yes you will need to tell the HDMI interface to do pixel repetition. But all TV video sources must do this, it has nothing to do with the Amiga resolution.

sigmason wrote:
Even if the frequency of the digital data stream amounts to NTSC video at 50 or 60Hz what about the resolution?

What about it?

sigmason wrote:
You are transmitting in digital per pixel. Usually most of the HDMI displays I've seen scale 480i to get it fullscreen because not only are they not 480i resolution

Yup. I've seen them range from 1920×1080 for full HD through 1366 x 768, 1280x720, 1024x1024 etc. etc. and they all have built in video scalers designed to upscale TV resolution to HD.

sigmason wrote:
but even if they are actually 640x480 they are missing 80 lines of information

What is missing 80 lines?

sigmason wrote:
So even 480i is actually stretching the timing in HDMI standard because it doesn't match one of the standard transmission resolutions so they resort to pixel repetition.

Ok. But pixel repetition is just supersampling right?

sigmason wrote:
The Amiga obviously can do more modes then that (even if he drops the modes you mention.)

But does it matter as long as you are sampling at the right rate? I don't think it does. They are all pixel compatible with a 28MHz pixel clock.

sigmason wrote:
I see a problem. If the RGB data coming out of Lisa is 320x200 interlaced for instance and you want to double it think about the timing difference going into something designed to take in 480i and getting something a division of 480i.  If they sample more then one pixel in digital RGB for that input timing how do they compensate for the timing difference?

I do not understand what you mean by 'timing difference'? Surely the HDMI transmitter is a slave to the pixel clock coming from the Amiga?

sigmason wrote:
I don't see inputs specified for chips from Silicon Image or Analogix that are lower then 480i.

It doesn't matter. All Amiga modes (except Super72, Euro72, and multiscan) will be compatible with 480i (or 576i). The only problem might be interfield motion, but I dunno how progressive displays fix this.

sigmason wrote:
I don't see RAM in there.

It doesn't need any. It just reads the same data twice.

sigmason wrote:
how would it compensate for this difference if you use them to produce HDMI from the Amiga and the Amiga is not in NTSC HiRes Interlaced or 480i?

The same way a TV does at the moment, oversampling. The Amiga keeps the RGB output stable a set period of time per pixel. When resolution is 320 pixels wide this time is roughly 2x as long as when resolution is 640 pixels wide because the line time is fixed.

sigmason wrote:
If you think I'm wrong enlighten me.

It is obvious from your last post that you are technically literate. We just have different ideas about different bits, probably each equally wrong ;-)

[alexh]

Eriond wrote:
I still have a question that Stedy addressed briefly, and that is how to make a simple I2C bus.

Two (3?) wires to the parallel port.

Eriond wrote:
My idea is to clip on the ADV101 to get all neccessary video signals (RGB bus, H & V-sync)

The video DAC doesnt have H & V.

Eriond wrote:
 but I still need to get to Lisas 28M buffered clock. So, it seems I would need to clip over both chips...

Or have a couple of wires that plug into the 23-pin video connector on the back (for H & V). It's been done before, you could get some DCE Flickermagic's which did that instead of clipping to the Alice.

[alexh]

freqmax wrote:
I think there's a chance to make this work with some 5V->3.3V level converters and an smart HDMI transmitter.

HDMI transmitters should have 5v tolerant I/O? Strange just on the I2C... So yeah you'll need some resistors on the RGB+Sync

freqmax wrote:
The worry points are:
 * Resolution and timing match, esp for 320x256 modes etc..

There is a high possibility it will be fine. If the HDMI transmitter works for regular TV inputs (and I am sure it does) then it should work fine for Amiga.

freqmax wrote:
 * Need for a possible external line buffer.

What for?

freqmax wrote:
 * Digital audio interface.. don't think Paula has one?

No, but you can just wire Analog to your Amp?

freqmax wrote:
Anyone have a A1200 full schematic?

Are these not complete?

[alexh]

Stedy wrote:
If you have the SiI9134 datasheet, check if it supports RGB 8:8:8 input (24 bit video)?

It says on the front page it supports 36-bit.

Do not get your notation mixed up.

RGB 8:8:8 means 8-bits per colour component per sample.

WHEREAS

RGB 4:4:4 means equal sampling of all three components, i.e. no subsampling/decimation like Y'CbCr 4:2:2.

[alexh]

CarlK wrote:
So, you connect a PIC to listen on a free adress somwhere in the IO space, and...? Will the PIC be doing some intelligent handling of the I2C bus, or just "reformat" it?

You don't really need a PIC to start with. Just two (3?) wires direct to the parallel port.

CarlK wrote:
:idea: Maybe, the PIC could listen to the commands given to LISA and automatically adjust the HDMI TX... hmmm?

I think you are going a little overcomplicated. I recommend the K.I.S.S. method. Setup the I2C registers (with parallel port) for PAL/NTSC and later you can add automatic switching (if necessary, I am not convinced that one setup wouldnt work for everything).

Anyone got the I2C register list? Of any of these chips? Some sort of driver source?