chrisNova777 wrote:

CelGen wrote:

I was bored so I grabbed a mac and installed Adobe Photoshop 2.0.1 with one of my old serials. […]
Show full quote

I was bored so I grabbed a mac and installed Adobe Photoshop 2.0.1 with one of my old serials.

which mac did u do this on? macintosh plus ?? 😀

SE FDHD. Almost the same machine but newer and supports 1.44mb floppy disks. There is however a Plus sitting on my bench right now though in the process of being refurbished. I'll probably put Photoshop 1 and 0.6 on that. 😉

Tetrium wrote:

MX4000 are universal AGP while 9200 tended to be AGP 8x ones (won't work in AGP 2x slot)

Good to know! I've got a massive gap in my knowledge about the myriad of cards between my Radeon 8500 and my next purchase which was a PCIe Geforce 8500GT.

@alexanrs
I have a MX 4000 PCI and it's slower than a Kyro 2 😁

Not much, except that I've been lugging all those boxes of parts back up stairs for the time being as I needed the room down here quickly 🤣

brassicGamer wrote:

Tetrium wrote:

MX4000 are universal AGP while 9200 tended to be AGP 8x ones (won't work in AGP 2x slot)

Good to know! I've got a massive gap in my knowledge about the myriad of cards between my Radeon 8500 and my next purchase which was a PCIe Geforce 8500GT.

I think these would be excellent as testing cards and they may be slower than Kyro 2 (which was around GF2 or so in performance?) but they are probably still faster than TNT2 and probably work fine for any retro rig, provided the games aren't too demanding. R9200...I have a few of those myself, no idea what to do with those as I'd prefer R9600 or perhaps even FX5200 instead.

Went through some of my boxes and found Acorp ATC-1415 board to play with. It had some problems as it would run CPU's only at 5V and it wouldn't run an ISA video card. So I tried it with ISA sound card. It worked fine. I already removed the leaking barrel battery when I got it. But it wouldn't remember CMOS data even with soft reboots. So I wanted to try external battery made from 4xAA that was already installed in one 386 computer. Even went ahead on google and it seems 6V is also OK. Checked for + and installed it. Booted and nothing... removed the battery and still nothing. I guess I put a final nail in the coffin 🙁 Than I checked what is the voltage output from the battery pack and it was actually 6.55V. I guess this was all too much 😒

I'm wondering what actually broke? The BIOS chip itself or something else?

I only have one more Award bios chip on a working VLB/PCI combo motherboard and I just don't want to experiment with it 😊

I messed around both with my Tualeron 1.2GHz (on a GA-6VTXE-A - Apollo Pro 133T-based + 512MB PC133 SDRAM) dual-booting XP SP3 and 98 and my Spitfire Duron 950 (on a PCChips M810LU - SiS 730-based + 256MB PC133 SDRAM). I have heard that SP3 is much slower than XP SP1 and below (and that those are close to Win2k), but I didn't think the difference would be great and with 512MB I thought the Tualeron would've been fine, but the humbler Duron DESTROYS the Tualeron. The Duron system feels snappy even when browsing the web with Firefox 3.5, and it feels like I could get by using it as a main computer if I had to, but the Tualeron just feels sluggish all around. Actually I was just setting up the Tualeron, but its sluggishness surprised me so much that I had to power the Duron on just to confirm it felt faster.

Is SP3 alone such a resource hog? Can it be solved with tweaking (disabling services, etc.) or is XP SP3 just hopeless and I should just format it and use 2K or XP RTM instead?

Found manual,added missing jumpers and fired up Asus TX97 first with K5-PR100 then with K6-2 450.
K6-2... i've try different jumper configurations,but it runs at 240mhz max,K5 works at 100mhz fine.
pros: it's alive!
cons:3rd ram slot cracked at end and i see some solder traces nearby south bridge

also washed with soap Labway LWHA151A00 Yamaha YMF719E-S based and added missing jumpers,tested and sounds ok

I finished running memtest on my -almost- sixty dimm sticks. Now I'm running hdd regenerator on 35 ide hard drives to see if they are ok. This is a torture 😵

Not retro but it will be used with retro stuff.

Im encoding a new source for encoding VCD, DVD and other test clips. The clip I wanted to use was a 3840*2160 50 FPS high bitrate H265 transport stream from an Astra satellite test transmission. Not many mpeg encoders can handle transport streams or H265 so Im transcoding the file to a 25 FPS 3840*2160 H264 MP4 file.

"Placebo" preset and "Warning RF 0 is Lossless!" = Im glad I have 24 threads at my disposal, the CPUs get a bit toasty with 100% load!

Skyscraper wrote:

Not retro but it will be used with retro stuff. […]
Show full quote

Not retro but it will be used with retro stuff.

Im encoding a new source for encoding VCD, DVD and other test clips. The clip I wanted to use was a 3840*2160 50 FPS high bitrate H265 transport stream from an Astra satellite test transmission. Not many mpeg encoders can handle transport streams or H265 so Im transcoding the file to a 25 FPS 3840*2160 H264 MP4 file.

"Placebo" preset and "Warning RF 0 is Lossless!" = Im glad I have 24 threads at my disposal, the CPUs get a bit toasty with 100% load!

encoding.jpg

I'm surprised you're still using the CPU for such a mammothly big conversion task. We've had GPU-Assisted video converting programs for quite a few years now. nvidia CUDA is the fastest. There are many free programs that will accelerate transcoding with the GPU. Freemake is the one I use. When converting 1080p video, adding a video card to the mix reduces conversion times down to 1/4 of what it would be without one, and CPU usage remains low.

EDIT: Just so you know, handbrake does not support gpu assistance.

kithylin wrote:

Skyscraper wrote:

Not retro but it will be used with retro stuff. […]
Show full quote

Not retro but it will be used with retro stuff.

Im encoding a new source for encoding VCD, DVD and other test clips. The clip I wanted to use was a 3840*2160 50 FPS high bitrate H265 transport stream from an Astra satellite test transmission. Not many mpeg encoders can handle transport streams or H265 so Im transcoding the file to a 25 FPS 3840*2160 H264 MP4 file.

"Placebo" preset and "Warning RF 0 is Lossless!" = Im glad I have 24 threads at my disposal, the CPUs get a bit toasty with 100% load!

encoding.jpg

I'm surprised you're still using the CPU for such a mammothly big conversion task. We've had GPU-Assisted video converting programs for quite a few years now. nvidia CUDA is the fastest. There are many free programs that will accelerate transcoding with the GPU. Freemake is the one I use. When converting 1080p video, adding a video card to the mix reduces conversion times down to 1/4 of what it would be without one, and CPU usage remains low.

EDIT: Just so you know, handbrake does not support gpu assistance.

I only care about quality and not speed when I prepare a source for MPEG encoder tests. I have already done both a 25 FPS and a 50 FPS version as the clip is only 12 minuts long. Im now creating a 720*576 25 FPS AVI file ready for the MPEG2 encoder with Staxrip, Im using Lanczos4 resize and lossless codec settings.

I will also try feeding the MPEG encoders a high resolution source directly to see if that produces a different result compared to resizing and reencoding with Staxrip or VirtualDub using lossless codecs.

Tried this GPU encoding a few years ago and never actually see any speed up 😕 It was on Phenom X6 and Geforce GTX460 and tested with different "gpu accelerated" video encoders.

kixs wrote:

Went through some of my boxes and found Acorp ATC-1415 board to play with. It had some problems as it would run CPU's only at 5V and it wouldn't run an ISA video card. So I tried it with ISA sound card. It worked fine. I already removed the leaking barrel battery when I got it. But it wouldn't remember CMOS data even with soft reboots. So I wanted to try external battery made from 4xAA that was already installed in one 386 computer. Even went ahead on google and it seems 6V is also OK. Checked for + and installed it. Booted and nothing... removed the battery and still nothing. I guess I put a final nail in the coffin 🙁 Than I checked what is the voltage output from the battery pack and it was actually 6.55V. I guess this was all too much 😒

I'm wondering what actually broke? The BIOS chip itself or something else?

I only have one more Award bios chip on a working VLB/PCI combo motherboard and I just don't want to experiment with it 😊

I don't think +6V is enough to kill the mobo, IIRC those RTCs are rated up to +8V. Also the external battery input generally goes through a diode so the actual input voltage to the RTC should be closer to +5V than +6V, well within specs.

Whatever it was, it's dead now 😵 🙁

Also tested some video cards. 2x Voodoo 1, Geforce 3 Ti200, 2x Kyro 4500... all dead 😢

2x voodoo... they don't seem to put any video out - with passthrough cable or directly connected in 3dfx mode (TOMB3DEM).
GF3... it doesn't put any video out... but computer continues to boot
Kyro 4500 1... even computer won't boot
Kyro 4500 2... it boots, screen flashes. When the picture is on, it's really fuzzy.

Not sure if this is my "lucky" day. So I'd better stop testing 🤣

Sorted out a lot of my hardware:
http://imgur.com/XVbU4JZ

Also tested some video cards and other hardware.

These seem dead:
http://imgur.com/GizCk8R

kixs wrote:

Tried this GPU encoding a few years ago and never actually see any speed up 😕 It was on Phenom X6 and Geforce GTX460 and tested with different "gpu accelerated" video encoders.

One part of getting speed up I find, is having the source file on a fast enough storage medium (SSD Raid-0 or something like that) and able to fed it in to the editor fast enough. If I have a video file on my mechanical drives, and try gpu editing, it's almost no faster than straight cpu. But if I have source file on my SSD's and feed it in to editor, gpu acceleration sarts getting down in to the 1/2 times, sometimes 1/4 of cpu speed times.

That's just my personal experience though.

I'm quite proud of myself for mostly repairing a broken C64 that I recently acquired. It was sold to me as broken and sure enough it did not power on when I tried. After opening it and cleaning the board up (a lot) I traced the main issue to the fuse connector no longer properly connecting to the board. There was so much gunk in that area of the board that I almost gave up hope but after scrubbing vigorously for a while I managed to get the board looking better. Must have been a drink spill because the top right corner of the board was covered in brown sticky gunk. As far as I can see the traces weren't too corroded luckily and since that part of the board is mostly to do with power the traces there are fairly think anyway. I tested the fuse and there was still continuity in the fuse itself, but there was no continuity when I tested between the points on the board that went through the fuse. I desoldered the fuse holder and after cleaning it and the board even more with isoprop I resoldered the connector and the machine booted. The screen was full of graphical corruption and the machine seemed to lock up after a few seconds however - but at least it was alive.

After trying a few different things I realised that if I removed the SID chip the machine booted normally again. Most games I tested (all except Turrican 2) work perfectly fine with the SID removed so I have to assume that the SID chip itself is at fault and is causing a short somewhere. Sadly the SID is probably the most expensive chip in the C64 but I see those SwinSID chips are not too expensive. I might end up importing one.

I finally got WinXP up and running on my main system! Man was that a pain in the ass.
Unfortunately I couldn't run any benchmarks with my GTX 970, (no driver support for XP), so I installed the next best card in my collection, a nice little EVGA GTX 680.

I'm gonna try and do a little more benchmarking before I plug my Win7 drive back in. So far I can't say I'm disappointed with the performance! 😀

kithylin wrote:

kixs wrote:

Tried this GPU encoding a few years ago and never actually see any speed up 😕 It was on Phenom X6 and Geforce GTX460 and tested with different "gpu accelerated" video encoders.

One part of getting speed up I find, is having the source file on a fast enough storage medium (SSD Raid-0 or something like that) and able to fed it in to the editor fast enough. If I have a video file on my mechanical drives, and try gpu editing, it's almost no faster than straight cpu. But if I have source file on my SSD's and feed it in to editor, gpu acceleration sarts getting down in to the 1/2 times, sometimes 1/4 of cpu speed times.

That's just my personal experience though.

At the time I just wanted to convert some DivX AVI files into MP4. So HDD speed isn't relevant. I was suspecting proper support for either DivX or MP4. Nowadays I just use RAW cpu power and VidCoder.

I decided to put together a 2nd Pentium Pro system, but I don't have an extra case. So I have parts laid out on a table with wires going everywhere. This has experience has convinced me to go buy a case later today.