Author Topic: Snap Vs Panorama - Refresh rate problem.  (Read 11107 times)

David Kiley

  • Full Member
  • ***
  • Posts: 131
  • Karma: +2/-0
    • View Profile
Snap Vs Panorama - Refresh rate problem.
« on: December 20, 2020, 08:32:25 am »
I'm running an old Radeon X300 graphics card, with a CRT monitor. I used snap since it seems to support refresh rates instead of Panorama, which doesn't seem to?
I found an article about Panorama below which seems to confirm it doesn't work with refresh rates.

But my problem is when I go into monitor settings using snap, the refresh rate supports up to 85hz and allows me to switch to it, but whenever the computer reboots it goes back to 60hz.
Is there a way to make it stay on 85hz - like a manual entry in config.sys or something?

Thanks for any advice.

Q13: What are the restrictions/disadvantages of Panorama VESA?
A13: Disadvantages of Panorama VESA:   
       * no 2D acceleration
       * no overlay support => slow fullscreen movies
       * no refresh rates => the driver is unusable for CRT monitors
       * requires CPU + memory
       * Widescreen is possible for ATOM BIOS and Intel only
          (read about Panorama widescreen activator)
       * No portrait mode (rotate by 90* while copying from shadow to screen buffer)
       * no DPMS support

Dave Yeo

  • Hero Member
  • *****
  • Posts: 4786
  • Karma: +99/-1
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #1 on: December 20, 2020, 08:50:32 am »
Try x:\snap\gamode,
Code: [Select]
gamode set <xres> <yres> <bits> <refresh> [head] [device]
   To set the default refresh rate for a specific resolution

Never used it but should work if you get the parameters right.

David Kiley

  • Full Member
  • ***
  • Posts: 131
  • Karma: +2/-0
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #2 on: December 20, 2020, 09:38:07 am »
Hmm. The command reports it sets the refresh rate - but when you reboot it doesn't actually stick.
I'm wondering if the command needs to run before the snap driver loads at boot - not sure how I would do that though. I tried making a startup.cmd file but that loads after the OS starts up.

Dave Yeo

  • Hero Member
  • *****
  • Posts: 4786
  • Karma: +99/-1
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #3 on: December 20, 2020, 09:49:41 am »
Sounds like a defect or regression. Try installing https://hobbes.nmsu.edu/download/pub/os2/system/drivers/video/snap-os2-3.1.8.zip and testing. Likely lots of complaining about older files and you have to enter the free registration.

David Kiley

  • Full Member
  • ***
  • Posts: 131
  • Karma: +2/-0
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #4 on: December 20, 2020, 10:21:56 am »
Wow you are the os/2 master :).
It worked.
So is it ok to run like that as a permanent change - meaning installing that didn't mess anything up with arcaos going forward?
(It did complain about old files like you said).

Thanks for the help!

Roderick Klein

  • Hero Member
  • *****
  • Posts: 655
  • Karma: +14/-0
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #5 on: December 20, 2020, 02:39:49 pm »
I'm running an old Radeon X300 graphics card, with a CRT monitor. I used snap since it seems to support refresh rates instead of Panorama, which doesn't seem to?
I found an article about Panorama below which seems to confirm it doesn't work with refresh rates.

But my problem is when I go into monitor settings using snap, the refresh rate supports up to 85hz and allows me to switch to it, but whenever the computer reboots it goes back to 60hz.
Is there a way to make it stay on 85hz - like a manual entry in config.sys or something?

Thanks for any advice.

Q13: What are the restrictions/disadvantages of Panorama VESA?
A13: Disadvantages of Panorama VESA:   
       * no 2D acceleration
       * no overlay support => slow fullscreen movies
       * no refresh rates => the driver is unusable for CRT monitors
       * requires CPU + memory
       * Widescreen is possible for ATOM BIOS and Intel only
          (read about Panorama widescreen activator)
       * No portrait mode (rotate by 90* while copying from shadow to screen buffer)
       * no DPMS support

I assume you got this list from ecomstation.ru ? That list is extremely outdated.

Widescreen support is possible with ATI and Intel Bios'es no wide screeen enabler driver needed.
DPMS support with the latest version of Panorama is working.

If you are running a single CPU system then using an version of Scitech SNAP from hobbes is possible.
However please note that 2D accelration by using native video chipset support is not the holy grail alone for good video performance.

The so called MTRR registers need to be setup correctly. If this does not happen on multi CPU system the video performance can be implode...
Your video performance will feel like as if you are behind an XT with slow graphics and text scrolling in a CMD window will be extremely slow.
The current Arca Noae SNAP driver fixes this.

As for perfomrnace with full screen, the DIVE performance Panorama is good enough to play full screen HD video's coming from Youtube smoothly.
As this low end laptop has two cores.

Roderick Klein

Dave Yeo

  • Hero Member
  • *****
  • Posts: 4786
  • Karma: +99/-1
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #6 on: December 20, 2020, 06:52:41 pm »
As Roderick said, the biggest problem is that the old Snap wasn't SMP safe as the MTRR registers need to be set for each CPU. Some BIOS's would correctly set them up, others left it to the operating systems drivers. I take it you are using old hardware, so most likely they'll be stable and fast for you.
Really an issue should be opened at Arca Noae as it is clearly a regression to not save the refresh rate. Most of us are using digital monitors where refresh rate doesn't matter.

Roderick Klein

  • Hero Member
  • *****
  • Posts: 655
  • Karma: +14/-0
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #7 on: December 20, 2020, 06:58:43 pm »
As Roderick said, the biggest problem is that the old Snap wasn't SMP safe as the MTRR registers need to be set for each CPU. Some BIOS's would correctly set them up, others left it to the operating systems drivers. I take it you are using old hardware, so most likely they'll be stable and fast for you.
Really an issue should be opened at Arca Noae as it is clearly a regression to not save the refresh rate. Most of us are using digital monitors where refresh rate doesn't matter.

You can goto https://mantis.arcanoae.com. And when doing so do not forget at attach a testlog.
See this page where to get the testlog program.

https://www.arcanoae.com/wiki/producing-diagnostic-log-files/

Roderick

Dariusz Piatkowski

  • Hero Member
  • *****
  • Posts: 1317
  • Karma: +26/-0
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #8 on: December 20, 2020, 07:15:46 pm »
SNAP and SMP operations...quite possible, but depending on your hardware/software config you may end up with a slow video performance.

Case in point: 850 XT PE video card, non-AOS SNAP release would work quite fine here (accelerated, etc.) but attempting to boot with the full 6 cores of the AMD Phenom II X6 CPU would sause a WPS blue-screen hang.

AOS has released some SNAP updates, their latest ones work perfectly fine with my setup and finally allow the WPS to boot-up with the full 6-cores of the CPU:

Code: [Select]
Graphics device configuration:
  Manufacturer......... ATI
  Chipset.............. Radeon X850 Series
  Bus Type............. PCI Express
  Memory............... 24576 KB
  DAC.................. ATI Internal 24 bit DAC
  Clock................ ATI Internal Clock
  Memory Clock......... 400 MHz
  Default Memory Clock. 400 MHz
  Maximum Memory Clock. 400 MHz
  Driver Revision...... 3.2, Build 29
  Driver Build......... May  2 2017
  Certified Version.... Not Certified

David Kiley

  • Full Member
  • ***
  • Posts: 131
  • Karma: +2/-0
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #9 on: December 20, 2020, 07:30:58 pm »
I assume you got this list from ecomstation.ru ? That list is extremely outdated.
Yep, that's were I got it So does Panorama support the highest available refresh rate? (since when I installed with Panorama it doesn't list any refresh rate under the screen application).

I think the motherboard is semi modern (Gigabyte GA-EP35-DS3L). I used the older graphics card since I found it in a database for compatibility. Using a xeon dual core cpu with ACPI installed.
So should I use Panorama?

David Kiley

  • Full Member
  • ***
  • Posts: 131
  • Karma: +2/-0
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #10 on: December 20, 2020, 07:34:54 pm »
Also, context - I use a CRT monitor since modern LCD monitors give me headaches.
Not sure how long that strategy will work since CRT is quickly becoming too old and probably the capacitors will burst.
But so far so good. :)

Dave Yeo

  • Hero Member
  • *****
  • Posts: 4786
  • Karma: +99/-1
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #11 on: December 20, 2020, 07:44:13 pm »
As was said, it depends on the hardware and BIOS whether the old Snap will work for you. Panorama is newer and likely the developers haven't taken refresh rate into too much consideration as so few people are still using CRT monitors.
If you notice bad performance, desktop hangs or crashes, you might have to revert to the newer Snap. Personally, I think if using a video card that Snap supports and if it works well, Snap is the way to go. Testing is the only way to judge though.
Note that sysbench (available on Hobbes) will allow you to benchmark the video performance.

David Kiley

  • Full Member
  • ***
  • Posts: 131
  • Karma: +2/-0
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #12 on: December 20, 2020, 08:06:55 pm »
The so called MTRR registers need to be setup correctly. If this does not happen on multi CPU system the video performance can be implode...

What is the MTRR registers and what is the way to set it up to not implode?
Wow you guys really know your stuff on os/2. Feel like a complete newb, and I know allot about Linux so thought I had a first class nerd pass but alas I apparently have a way to go lol.
« Last Edit: December 20, 2020, 08:08:26 pm by David Kiley »

David Kiley

  • Full Member
  • ***
  • Posts: 131
  • Karma: +2/-0
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #13 on: December 20, 2020, 08:11:20 pm »
As was said, it depends on the hardware and BIOS whether the old Snap will work for you. Panorama is newer and likely the developers haven't taken refresh rate into too much consideration as so few people are still using CRT monitors.
If you notice bad performance, desktop hangs or crashes, you might have to revert to the newer Snap. Personally, I think if using a video card that Snap supports and if it works well, Snap is the way to go. Testing is the only way to judge though.
Note that sysbench (available on Hobbes) will allow you to benchmark the video performance.

Thanks.. Snap seems stable so I will go ahead with that. The performance seems OK.

Dave Yeo

  • Hero Member
  • *****
  • Posts: 4786
  • Karma: +99/-1
    • View Profile
Re: Snap Vs Panorama - Refresh rate problem.
« Reply #14 on: December 20, 2020, 08:40:00 pm »
The so called MTRR registers need to be setup correctly. If this does not happen on multi CPU system the video performance can be implode...

What is the MTRR registers and what is the way to set it up to not implode?
Wow you guys really know your stuff on os/2. Feel like a complete newb, and I know allot about Linux so thought I had a first class nerd pass but alas I apparently have a way to go lol.

Memory type range register, basically they control caching of memory ranges.
Here's an old article about setting them up in Linux, https://www.meduna.org/txt_mtrr_en.html on a single processor system. The thing is is that every CPU (core) has to be set up the same.
Here's the current status on Linux, mostly superseded. https://www.kernel.org/doc/html/latest/x86/mtrr.html