SGI: Hardware

Hi all ...Long time, no post...

Is there a precompiled list of modern day LCD flatscreen monitors that can handle SGIs output ?? If one does exist, here's two more that behave OK..


- LG Mod. L1942T 4:3 format ...13w3 to HD15 VGA...on an Indy.............works perfectly

- Samsung SyncMaster Mod. 2043nw 16:9 format ....13w3 to HD15... on an Indy...also works perfectly (narrow 4:3 image)

Now I can finally retire the big power suckin' Sony

I think GeneratriX would add these into:
viewtopic.php?f=3&t=16720341 )

concerning analog-connected LCD flatscreens: the proper VFO-configuration-file ist most important for image quality. I for myself bought another 13W3-HD15 adapter in the belief, the poor image quality was due to poor cable quality, when I became aware of this thread:
viewtopic.php?f=3&t=16725909

the cable I had to connect to the dell was OK, SGIs predefined 1920x1200 VFO was not that good for the DELL screen

Done!, with a few more aditions too!
All the best!,
Diego

Diego, do you have a VFO for Dell's UltraSharp U2412M LCD? It's listed in the aggregator...

Not really. Never tried it myself. Sorry.

In fact I was thinking if could be possible to compile some 1920x1080 VFO for Octane / ESI for another flat panel, since I don't know if the range of resolutions from "IMPACTSR" on IP30 would support such thing.

My favorite kind of VFO's are the ones that other people make for me...

LOL! ...kinda like cooking meat. It always tastes better if you don't eat by yourself all that smoke around the grill!

Anyway I compiled a few VFO's in the past for IP32, a platform where I happened to feel slightly more at home than with IP30 (spent a lot more of time with it) anyway... I would not try it again without further investigations. You know; there is something very evil around bad formed VFO's, it is pretty much like a Lemarchand's box!

Somebody around here must have compiled a VFO for Dell's U2412M, and one of these days I'm going to find him!

And actually, my favorite kind of everything is OP's (other people's)...

I guess you already tried here? --->

Any experience with IPS SOG-capable displays ?

Yeah, he's using some funky custom cable that I haven't had a chance to figure out yet...

LOL! ...the hairy cable should not make any difference; I guess the 13W3 connectors are a bit scarce now, I could not find spares now over here, but looks like he used the D-Sub from some VGA cable to bridge either the original SGI cable or the 13W3 connector in it.

BTW: anyone knows if it is possible 1920x1080 @ 60 Hz for Octane / MardiGras stuff?

Yeah as I mentioned at the other thread I'm trying to run my Dell through a dual channel display, and the dual channel display outputs are DVI...

I can never remember what board the MardiGras is (probably because I never had one), is it MXI or SI or SSI? Never had any of those either...

Oh, sorry; my experiencies with workstations from SGI only ranged between 13W3 and D-Sub equipped graphic stuff!




(Source: http://www.sgistuff.net/hardware/systems/octane.html )


But since they are all of them reported by 'hinv' into a different fashion...


(Source: http://techpubs.sgi.com/library/tpl/cgi ... G/apa.html )


So, my setup has an SE and an SE+TRAM.

Nice short and useful summary for my memory!

The offer still stands.....

Is there an EDID extractor that works for IRIX? Me doubts it. I'll look and report back status...

You may have to get your hands wet in the Windows urinal

For IRIX? Not that I know of, but as hamei so lyrically suggests, they do (freely) exist for OSX, Windows or Linux.

If you don't have something that'd run one of the EDID extractors, perhaps a friend with a suitable laptop could offer a few minutes of assistance. My experience so far has been with OSX; extracting EDID info with SwitchResX was a very short and simple process.

Or if you'd like to leave an indelible mark on the IRIX world, the source code for read-edid is available here . I'm not source-code or compiler qualified, but have seen enough mentions of other hardware oriented open-source tools get shot down in flames that I won't be surprised to hear the source isn't IRIX or SGI hardware compatible.

Depending on your monitor it might prove to be a tight fit. You might also want to obtain EDID info for your monitor - it might prove to be helpful.

The maximum pixel clock for MGRAS graphics is 160MHz. I tried to create a 1920x1080_60 MGRAS format using the VFC block sync script with the MGRAS chip and board definition files. The default timing used by VFC at 1920x1080_60 exceeded the MGRAS 160MHz pixel clock ceiling and produced a "no format created" error:

Not knowing anything specific about your monitor, I used CVT to generate a generic 1920x1080_60 modeline with VESA standard timing (likely very similar to what the VFC Block Sync tool produces). That resulted a too-high-for-MGRAS pixel clock of 173MHz: If all else fails, and *if* your monitor will support a format with reduced blanking , you may still be able to get around the 160MHz pixel clock limitation by creating a format with reduced blanking. Here's another 1920x1080_60 CVT modeline, asking for reduced blanking lowers the pixel clock to 138.5MHz: VFOs with reduced blanking haven't always been successful, so perhaps the EDID info will show that right out of the box your monitor uses a pixel clock lower than 160MHz at 1920x1080_60.

This

sounds like sacrificing Clytemnestra before we even leave home ...

That's the 8086 assembler and loader, the code could possibly be modded to use the equivalent that comes with MIPSPro...