Prevue Guide Forums

AriX wrote:I'm seeing some other interesting stuff now - when an external video source is plugged into the genlock, keys 5-8 have an effect. Key 5 reads "Ext. Video Only", and perhaps as expected clears all video off the Amiga's screen in favor of the stuff behind it. If the external source is unplugged, the Amiga's output will automatically pop back up. Key 6 reads "Computer Only", meaning, of course, show only the Amiga's output. Key 7, "Overlay Ext. Video", does the same thing as the "n" key, but in the menu. Lastly, key 9, labeled "Negative Video", swaps layers - the genlock input suddenly appears over the Amiga's output instead of vice versa. The odd thing is that these keys seem to do more than just tell the genlock what to do - when I plug the serial in my computer and use a program called "Serial Tools" to view its state, it is clear that pressing key 5 lights up both CTS and DSR. 6 lights up CTS only, and 7 and 8 light up neither. Keys 1-4, (Silence, Left, Right, and Background respectively) do not have an effect on CTS or DSR.

Question is, why would the Amiga be sending this information through its serial port?

5-8 Makes sense, as the genlock is being controlled (probably) via the parallel port as it has no external controls for the kind of thing you describe so the software must do it.

I'm interested in the CTS/DSR stuff because it's like the bitbanging we already discussed at length, but I don't think it can do anything, as there doesn't appear to be the hardware to handle it on the demod card. I guess maybe it was for future use?

BTW I looked into the audio modulation and have drawn a blank.

tin wrote:

AriX wrote:I'm having some trouble sending data to the Amiga over serial. My computer is hooked up to its serial port, but it's not recognizing data. The DATA counter goes up, but it doesn't see any commands. :/

You must be sending it wrong then DATA counter indicates valid serial information coming in, Commands indicates something valid has been acted upon.

Yeah, I'm definitely doing something wrong. Here are your photos, by the way: cards/new/

Update: Seems like Serial Tools was having issues - I booted into Windows and used HyperTerminal, and it received my 55AA2A0094 just fine. Looks like I've got a little more work to do.

AriX wrote:Here are your photos, by the way: cards/new/

Sorry to be a pain, but are you able to get ones that are a bit more in focus? I'm trying to trace the tracks and it's a bit difficult on blurry photos

tin wrote:

AriX wrote:Here are your photos, by the way: cards/new/

Sorry to be a pain, but are you able to get ones that are a bit more in focus? I'm trying to trace the tracks and it's a bit difficult on blurry photos

No problem at all, sorry about that. I was trying to take them very quickly. It could be a little bit though. In any case, I got the serial working from my Mac - I really don't know why it wasn't working in Serial Tools, but I used a little serial UDP listener I wrote a while ago (that didn't work on the Atari) and it works great.

In looking at the photo of the bottom genlock output, finding the video slot pinout, and deducing that both pins of that connector won't be ground, it almost looks like that output is using the digital B line (as would almost never be used for output to a CGA monitor, along with R, G, and I). Counting in one direction (since I'm not sure which side is pin 1), the two pins go to pins 12 and 24, and counting the other direction, it uses pins 25 and 13. Pins 12 and 24 are both ground, ruling out that option. Pin 25, however, is listed as "B3 = DB", and pin 13 is ground, making me think this is the case. I must be missing something, though, because:

A) TTL RGB as used in CGA is bi-level - on or off, especially in this case where only one of the four lines is being used.
B) Where the hell is the sync coming from for that output? CGA-style RGBI normally has separated h/v sync, so that picture should be garbled!

Am I missing something here? Have I somehow traced back to the wrong pins? I don't see anywhere on the video slot where the mono video is present, so the only other real option is that somehow I mistraced the lines, and it's using one of the three analog outputs plus composite sync. At first I thought I had a sort of answer, as a blue-only feature is common on professional equipment for calibration purposes (when I worked in news production, every one of our major production monitors had a blue-only switch). However, as mentioned, it can't be the TTL B line by itself or there would be no sync and no variation in signal level (either fully on or fully off).

Nice detective work!!

I did almost the same detective work but don't have the video background you have so I'll say what I know and then leave it to you

The pins are actually 12 (component side) and 13 (solder side) and the rear half of the connector is the extended video slot on the page you linked (except the page lists the A4000 extended video slot so it shows 54 pins rather than 36 so I guess that's what threw you off).

12 is ground (actually analogue ground as they are different in the video slot)
13 is listed as "sync-on-green indicator"

Now. as I say I'm no video hardware expert at all, but something rings a bell about Sync-on-green sometimes being the same as composite, or maybe b/w, or maybe sync IS b/w or something like that. I'm really not sure. Maybe you can shed more light.

For ref, this is the version of the A4000 pinout I found that tells a slightly better story. I know there's some previously reserved pins changed from the 4-bit to 8-bit versions and of course therefore there are more pins in the A4000, but they are pin compatible.

Ok, well I'm not a hardware guru so I wasn't sure if I'd traced them properly, but I definitely used the left-hand side of the page I linked to (which matches the number of pins on the card if you don't treat the pins as doublesided). I based that on counting the pins and finding that there were 36 total, and assuming that the "standard" video slot was 2000/3000 and the "extended" slot was 4000. Obviously I was wrong, as my findings didn't match up. SOG would kinda make sense if that is the case, as it's basically the green component combined with the h/v sync, which should display as mono video on any NTSC-capable monitor. I spent about 30 minutes researching this and it seems that detailed video slot information is scarce. A good way to test and confirm would be for Ari to make a few screenshots from that output (captures if he has the capability) with different types of content on the screen, and then taking similar emulation screenshots and stripping the R and B channels. If this is correct, then the resulting images would be nearly identical to the signal that output carries. Given this, though, I'm thinking that the real purpose of this signal wasn't for any type of monitoring, but rather for upstream video equipment to lock onto it (although the same information is present in a composite signal, lots of video equipment has separate genlock inputs to help assure that all video sources were properly timed). Odd, though, since the sync coming out of the Amiga is going to depend on the input video (in fact, running video through an Amiga genlock slaves the Amiga to the input timing - I remember back in the day many Amiga crashes due to using a SVHS VCR without TBC as input and searching through the tape too much).

I also noticed that out of the two variants of UVSG genlocks that I've seen, Ari's is more like the Sneak Prevue than the Prevue Guide genlock, which does not have this third BNC. Either Ari's unit was never a Prevue Guide machine, or UVSG had a habit of repurposing machines between the two projects without rebranding them. Wouldn't surprise me - I own an A3000 that AFAIK has always run AmigaOS, but yet on the back it had a sticker (that fell off, but I kept) labeling it as an A3000UX. The case is painted, though, so I have no idea what the actual badge says.

LocalH wrote:Ok, well I'm not a hardware guru so I wasn't sure if I'd traced them properly, but I definitely used the left-hand side of the page I linked to (which matches the number of pins on the card if you don't treat the pins as doublesided). I based that on counting the pins and finding that there were 36 total, and assuming that the "standard" video slot was 2000/3000 and the "extended" slot was 4000. Obviously I was wrong, as my findings didn't match up. SOG would kinda make sense if that is the case, as it's basically the green component combined with the h/v sync, which should display as mono video on any NTSC-capable monitor. I spent about 30 minutes researching this and it seems that detailed video slot information is scarce. A good way to test and confirm would be for Ari to make a few screenshots from that output (captures if he has the capability) with different types of content on the screen, and then taking similar emulation screenshots and stripping the R and B channels. If this is correct, then the resulting images would be nearly identical to the signal that output carries. Given this, though, I'm thinking that the real purpose of this signal wasn't for any type of monitoring, but rather for upstream video equipment to lock onto it (although the same information is present in a composite signal, lots of video equipment has separate genlock inputs to help assure that all video sources were properly timed). Odd, though, since the sync coming out of the Amiga is going to depend on the input video (in fact, running video through an Amiga genlock slaves the Amiga to the input timing - I remember back in the day many Amiga crashes due to using a SVHS VCR without TBC as input and searching through the tape too much).

Interesting. Sounds like a good idea. I need to get my digital capture stick out so that I can take some screenshots.

I also noticed that out of the two variants of UVSG genlocks that I've seen, Ari's is more like the Sneak Prevue than the Prevue Guide genlock, which does not have this third BNC. Either Ari's unit was never a Prevue Guide machine, or UVSG had a habit of repurposing machines between the two projects without rebranding them. Wouldn't surprise me - I own an A3000 that AFAIK has always run AmigaOS, but yet on the back it had a sticker (that fell off, but I kept) labeling it as an A3000UX. The case is painted, though, so I have no idea what the actual badge says.

Interesting. I'm not totally surprised. I think this may be somewhat of a frankenstein machine - either they were out of genlocks at one point and ended up using a Sneak Prevue one, or someone had their hands on multiple Prevue machines after they were discontinued, and mismatched some of the parts - my first clue is that I have labels from both types of machines - I have a big Sneak Prevue logo, but I also have a Prevue Channel keyboard flag and a Power/Control Data sticker. One thing to note: this Amiga 2000 was never used for the Prevue Channel. The stickers, cards, software, and ROM were lifted from an original Prevue machine and transferred to a regular one in 2004 or so, because someone bought the Prevue machine with the intention of using the Amiga and sold the Prevue stuff itself separately. This is why the stickers are not well attached to the machine, and it is missing some elements (select code on the back in sharpie, Prevue Networks asset tag). These parts have changed hands at least 4 times from what I have seen.

I've been pretty busy, and I won't have much time to focus on this project for at least another few weeks, but I did get around to taking (hopefully) better pictures of the Amiga cards. I did it pretty quickly, so if you need more/better, just ask (again :p). They're easy for me to access now, at least.

http://prevueguide.com/AmigaCards2.zip

Excellent, I'll probably get time to have a good look at them maybe wednesday