jheronimus wrote on 2020-04-15, 13:46:

Can you play MP3 music without it being choppy?

Kind of a weird use case for me 😁 Then again, I once listened to a Pink Floyd album on CD using a 486 computer — for no apparent reason.

Well - back in the day, when I last time played with a 5x86, the MP3 playback in WinAmp was the indicator for me - at 133MHz choppy music, at 160 no hick-ups 🤣

Me, too. Especially in early games where the textures weren't hi-res. Filtering turns these into a blurry mess.

imi wrote on 2020-04-15, 21:16:

that sounds interesting, I really prefer the sharp pixelated rendering over early texture filtering in a lot of games.

I further enhanced the experience by altering the filtering and setting particles to square in the console, I will have screenshots soon..

appiah4 wrote on 2020-04-16, 07:15:

imi wrote on 2020-04-15, 21:16:

that sounds interesting, I really prefer the sharp pixelated rendering over early texture filtering in a lot of games.

I further enhanced the experience by altering the filtering and setting particles to square in the console, I will have screenshots soon..

Here they are, GLQuake on ATI 3DRage Pro (with some extra rendering tricks: gl_texturemode GL_NEAREST_MIPMAP_LINEAR, r_particles 2)

Most visually appealing GLQuake experience I've had, personally..

Whoa, that really looks very much like software rendering.

looks great :3

some textures still seem to get filtered though apparently?
wall in the first screenshot and the floor texture.
at the door you can see the side walls being filtered while the front facing ones and the door aren't.

very weird, can't you just turn off filtering in glquake altogether for all cards via console commands?

Filtering can not be completely disabled, but gl_texturemode GL_NEAREST_MIPMAP_LINEAR changes filtering mode to point sampled instead of bi/trilinear which makes it as software-like as possible.. By default 3D Rage Pro can only do bilinear filtering to non-alpha channel textures anyway, but this makes things even more software render like..

flupke11 wrote on 2020-04-12, 15:19:

I gave the P3TDE6 a new home in an Antec/Chieftec full tower from 2001. […]
Show full quote

I gave the P3TDE6 a new home in an Antec/Chieftec full tower from 2001.

P3TDE.JPG

Specs:
Supermicro P3TDE6, dual Tualatin 1,4GHz, 2GB PC133-ECC
USB: VIA 2.0 on PCI
Sound: Hercules Fortissimi II on PCI
Disks: 9.1 Ultrastar 2*36 Seagate Ultra-SCSI, 2*150 WD Raptor in disk cage and 1*40GB Intel SSD on 3Ware 9550SX-8LP (PCI-X 66MHz 64-bit)
Optical: 2* Panasonic DVD-RAM on SCSI, 1 Plextor PX-116A on IDE
Legacy: 5'25" and 3'5" floppy's
Removable: 1 ZIP250
OS: WinXP

Purpose: SCSI test bed, Gotek base

edit: replaced SCSI disk

I'd love one of those full towers! Sadly I dont have the room 🙁

FazzaGBR wrote on 2020-04-16, 11:29:

I'd love one of those full towers! Sadly I dont have the room 🙁

The idea is to have all my ATX Dual CPU systems in such Chieftec/Antec towers, which I got cheap throughout the years (and in various states of abuse). I'm a big fan of Chenbro cases, and a fellow Vogon'er has his favourite systems in such cases, but sadly they are hard to come by where I live (and frightfully expensive otherwise). So these are a cheap man's second best.

So at the moment the projects-in-waiting to fill these boxes are:

Dual Socket 8 PPro Overdrive
Dual Slot II Xeon
Dual Socket Celeron (BP6 😀)
Dual Socket 604

I'll keep the Dual Slot 1 Compaq AP550 intact, no need to change cases as that hull is already built to withstand a nuclear winter.

My Tulip AT Compact(s) (NEAT chipset 286) 3 is/are proving to be a nice project. I own two of these systems, both with their own problems. I decided to combine them into one.

One mainboard is 'dead', it does not POST. The other seemed to be okay, but some tantals exploded when powering it up (I posted before)... After replacing them all seemed okay.

I could not get a serial mouse working... I tried an external loopback plug, but tests failed. After some visual inspection and datasheet reading, I think the DS1488N chips are fried. They have some visible damage on top (a little bit visible in the picture), possibly/maybe as a result of the tantals blowing up? Or they were the cause... We'll find out when I replace them. Desoldered them today and put in sockets... Spares should be here on saturday, I found a cheap local seller.

Tested the motherboards and PSUs I picked up this week. I was hoping maybe one would work, but all four boards do, so do the CPUs on them and so do the power supplies - which is relevant as one board is a Dell-modified Intel SE440BX-3 with non-ATX pinout and one other is a Compaq board from 1999 with a 24p ATX-ish connector. Now the usual problem: what on earth to do with them. I might well use the SE440BX myself, but I'm overflowing with P4 and Athlon64 systems I have no use for and no one is interested in buying (hell, I can't even give the things away). Taking to the recycler would be the logical choice, but I can't bring myself to do that - yet.

Tested a 386 motherboard I bought ages ago as working but never actually confirmed for myself. Good news is that it does in fact work! Now if only I had something to put it in.

Finally got the new laser for my PlayStation SCPH-1002 calibrated correctly.

Hint: don't follow any of the guides trying to tell you about the correct voltages. And use an audio CD for calibration, not a game disc! Adjust BIAS to the lowest voltage at which a disc is still detected. Then start playing the audio CD. Now - while the audio CD is playing - adjust GAIN to find the left-most and right-most point at which the audio is no longer playing correctly. Then set it to the middle position between the two and you are done. Well, almost. The same can be done on the laser itself, but that should be correctly adjusted normally already. Though it wasn't on mine. It was set to 14mV, but I determined it should be 16mV.

Final test must always be done with the PlayStation completely assembled, as any sort of scattered light will negatively affect the laser's ability to read the disc.

More hardware accelerated 3D testing with the ATI 3D Rage Pro, this time Quake 2 OpenGL:

Again, even at default settings the general texture filtering smear my brain associates with the Quake 1/2 engines is not present here, probably because 3D Rage Pro can't filter alpha blended textures.. Regardless, without even setting the filtering to NEAREST_MIMPMAP_LINEAR (so at GL_LINEAR_MIPMAP_NEAREST by default) it looks a lot like software rendering. Blessing? Curse? You decide.

It's more because GLQuake/Q2's accelerated look in 97-98 was mostly associated with 3dfx in which they've got a smeary filter with default high gamma (so you can have less contrast for even more obvious smeary), whereas ATI here is one of the many that were considered "too unplayably dark !!!1" in which that it's technically correctly rendered.

Oiled the fan on my TI4400 AGP as the fan was stiff and barely moved. Now it runs faster so that it doesn't run hot all the time. Fan is still a bit noisy, but it runs.

Also did the bearing on my modern desktop since the fan's bearing went out and after oiling that, it works.

Well, it has been a challenging day for me. My cheap, $8 Radio Shack dual mode (15/30w) soldering iron gave up the ghost - I was about to use it to desolder a leaking capacitor off my t5720 motherboard (one that, I might add, had RoHS solder which makes it a pain to work with) so I can replace it with better caps from Digikey.

Then I realized that I had the iron sitting for at least 7 minutes, but it's not giving out much power/heat - the infrared/laser thermometer that I use to spot check the temp did not show anything above 65 Celsius along the device. Honestly, good riddance. The tip was bent (and really never steady), the cheap plastic handle did not protect against the heat from the element right in front...and the cord was too short and unwieldy. Time to buy a new one - probably a 70 watt Weller digital soldering station so I can get that RoHS garbage off. Amazon is quoting shipping to NYC for mid-May, which is...unfortunate.

The replacement t5720 motherboard I bought had a bad northbridge and rejected all the DDR RAM at my disposal (why does it say my memory bus is 100Mhz when it should be 266, 333 or 400?) and had been sent back for a refund. In the meantime I ordered a replacement for that replacement. The motherboard re-capping for the original broken board was really to be done as a learning exercise.

On other (happier) news, got a bunch of 4GB MicroSD cards that I can use to do some experimental OS installs on the good old Thinkpad 560E. Right now, it's QNX 4.2, OS2 Warp 4, Windows NT 3.51, Windows 95B, Windows 98, and possibly AIX 1.3 for IBM Thinkpads. Trying to see if it's possible to source a cheap PCMCIA-CDROM drive so I can do some CD-based stuff like NT 4.0 or NEXTStep.

Man, that has to hurt big time. I had a RadioShack soldering iron or might still have it and it takes forever to heat up. My parents got me one for Christmas and it works a lot better than the old irons I had that either took forever to heat up or the tips kept falling off all the time. I like this one better, but I set it to 250C to tin and clean the tip off, then turn it to 350C for soldering and takes 5-10 seconds to heat up, which is really cool. I, myself, need a hot air station so that I can remove and install SMD/SMT QFP video chips on a motherboard that I'm planning on upgrading the chip from a 5428 to a 5429 for better DOS acceleration.

Just catching up on this week's retro activity - mine happened earlier in the week and not exactly today.

I purchased a brand new sealed NOS I/O chassis plate for my Chaintech 7VJL Apogee KT333 motherboard and fitted it. Getting the motherboard out, slotting on the new I/O plate and screwing it all back in without removing any components was a bit of a task but got there in the end. Finally, I just gave the machine a quick dust out.

Recapped my Gigabyte GA-5AX.