AMD Catalyst (Modified Build)

Yeah, have to get few people with HD7 at some point to get this topic back on track, but I doubt the hardware has changed that much yet for it to make difference.

As for the cleaner yeah it's still fully functional cladly not that much has changed, but by now I believe some other cleaning systems has already
implemented the only few new ideas I made inside that cleaner quite sometime ago like GUID detection > clean.. and so on..

To be honest just so little time these days, and maybe a bit low interest too on one point as had to work pretty near around the clock still losing stuff.
Hardly no equipment to test against theories I had/have makes incredible hard to mod where I used to have at least barebone testing lab of my own.

 

it might be a waste of time to mod NOW until we see the effect of dropping whql+old cards has (hopefully a good one)

 

actually this line from AMD release notes from windows 8 drivers pretty well sums it up:

heh.. I honestly believe optimization would be really good, for the moment... There's absolutely nothing what in Windows 8 could slow down DirectX 9 it's
the driver branch they have half build whole last year. Since what they keep burying modders usually dig the coffin up and only then big companies move a bit to right direction.

 

well i'm ignoring win8 ones, i mean for trying to performance/stability mod xp/vista/7 drivers

 

Ok, I think we finally actually have something going on here. Seems the XP driver as in DX9 beta was just released and that should work as very good base driver for main builds. Although, Crossfire driver has to wait.

I updated the Clean-up and verified against old data to new data, included couple missed new AMD paths and OpenCL stuff.

Genetix Cleaner v5.01 - Stand-Alone ATI/NVIDIA clean-up system - [06.06.2012]:

* Included OpenCL new data.
* Included missed data storage folder.
* Verified new content against old one to be accurate.

x86 / 32-bit - x64 / 64-bit:
Mirror 01 [ZippyShare] - Mirror 02 [MediaFire] - Mirror 03 [Multiupload]
MD5: cfd8f33aa3b195de16b532e4212119db

And have to see, if I can get the drivers back at some point I'm hoping this new one reserves as suitable beta.

 

You back!
Thanks TwL!!! You always deliver!!
I was getting sick of the "other cleaner" if you know which one i mean. (Wasn't DriverSweeper.)

 

Yeah, looks like I need to build new driver on some point I am working one, atm. Needs day or two testing, but looks pretty solid build. The new
Beta(not now referring the 22. Beta which is pretty good) actually is build for laptops, btw, which is kinda weird one, but all good except the CCC2 sucks
in it. However, seems yet again we have unfunctional AF and AA methods from AMD driver team. So, I guess have to at least do something about that.

As for DriverSweeper well the difference there is (as you know) I build the cleaner for myself not to people here or not for general use, but still added it
online since it's probably only thing out there still, going through GUID detect/identify > removal progress. Software like DriverSweeper is again
build to the people and not for the developer only use.



Cleanup 5.01 issue - Sidebar lock folder<->log file deletion until rebooted:
(have to fix that sidebar lock down on end of cleaning by remove relaunch. heck it relaunches anyway on next restart and needs driver anyway after
cleanup. More of an formality, but the darn thing locks the log file until rebooted or sidebar is closed.)

 

Cool! Can't wait for it.
So far, 12.6 beta is being wonderful for me. Except for MPEG2 deinterlacing problems. It doesn't do it properly and the driver tend to crash while at it. It just sucks (happening since 12.3 i think). This is while using HW acceleration of course. (Just submited a survey thing through the new survey page AMD Issue Reporting Form )
But in games, it just rock the house.

As for the issue you mentioning with the cleaner. I still haven't used it (waiting till you release a the driver) but, for me it won't be a problem, as i don't use sidebar.
Thank you again!


PS:
If helps you for anything in case you want to fix the MPEG2 HW acceleration issue, the file crashing is aticaldd.dll
12.1 worked perfectly fine.

 

There most be something wrong on your player video output rendering settings, if that would be the issue file. I mean that's the main process file not even execute direction or detection file. Check that you are using stock EVR should fix that although I know there's a lot of issues on these AMD decoders they are just plain bad I personally use the ffdshow64 experimental builds with all audio disabled on ddshow audio side > Configured AC3Filter (new versions) instead of using the AMD stuff to anything else at least than WMV / MP4 / MKV compressions.

 

AVC, VC1, MPEG4 and the like are no Issue.
But, MPEG2 is.
I use ArcSoft Total Media Theater for DVD playback (with this one is the crash), but with Windows Media Player, Power DVD 12 and Splash Pro EX the video gets deinterleaced incorrectly (bad field order, so it stutters like ****ing hell) In all of them I have to disable hardware acceleration.

And here is the issue http://www.arcsoft.com/forum/forum_...=1&title=amd-driver-crash-regular-dvds-521119

It has been confirmed that is a driver problem, because with 12.1,2 and 3 it was working fine.

 

this is with automatic deinterlacing set in CCC right? what about pulldown?

 

Yes, automatic deinterlacing (although, manually changing it does nothing, still bad deinterlace) And Pulldown activated too (again, does nothing if I change it)

 

GG.

I always use your cleaner before trying Catalyst sets.
"Nuke it from orbit; just to be sure.", they said.
Just so that I'm sure if there's anything wrong with a set then that means it's a bug on that specific set, not a carryover from a previous installation.

Thanks, TwL!

 

Well, don't know about the MPEG2 issue. Sounds some kind of buffering/streaming read issue as for pulldown yeah there's punch of things still not taking any effect and could maybe changed on registry, but catalyst won't change them by changing them in the app.

Have to test that with something. Any chance of copying me exact audio/video format/codecs used?

 

Spoiler

ID : 224 (0xE0)
Format : MPEG Video
Format version : Version 2
Format profile : Main@Main
Format settings, BVOP : Yes
Format settings, Matrix : Custom
Duration : 14mn 11s
Bit rate mode : Variable
Bit rate : 9 215 Kbps
Maximum bit rate : 9 800 Kbps
Width : 720 pixels
Height : 480 pixels
Display aspect ratio : 4:3
Frame rate : 29.970 fps
Standard : NTSC
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Interlaced
Scan order : Top Field First
Compression mode : Lossy
Bits/(Pixel*Frame) : 0.890
Stream size : 936 MiB (94%)

Can give you a video sample if you need one.
As for the exact MPEG2 encoder, in this case is Cinema Craft Encoder SP3
But, it crashes with all kinds of MPEG2 encodes. (MainConcept, FFMPEG, etc)

 

Won't check in to it today as have no time today, and no sample won't help. I need longer stream to test as, if the problem is where I think it is there's probably no driver than can fix it.

-update-

OK, added an experimental build online. This build is meant for single card usage for now new or old. DX10-11 will take a hit. Build seems quite stable
tested only few days for it. There's multiple areas to fix in 1 single driver currently we have video bugs, we have AA bugs, we have everything it seems
that we used to have broken and I am working on it. Bugs will be hunted down 1 by 1 and fixed as I used to fix them (, if possible).

I named the build as v5.15 Experimental as it's my currently running system work build doesn't mean it doesn't work for all systems. It comes seconds
faster loading of multiple things from AMD driver compared, but the processing is still off and no where near what we had year ago, so, there's a lot of work to do.

Future Buids:

Builds will go as follows:

v5.16: Video progressing improvements -> skip correction, AMD noise crap total removal (They actually ADD the noise to your videos not fix them mostly).
v5.17: Eyefinity / multi-display setup fixes.
v5.18: CFX/TFX processing fixes.
v5.19: Split x86/x64 systems -> general update -> Further 3D testing through DX layers.
v5.20: Final Builds.

This might take 2-4 weeks to go through. Hopefully I have time to build them I have a new apartment coming, so, that might cause at least week
cut making this 4-5 weeks minimum to go through and by that time we should have decently working DX10-11 build online to clean off for Final build.


-update-

bugs at v5.15:

* Return from power save mode directly to 3D enviroment might crash display driver.

 

Can I safely use these drivers with a 6950 CFX setup?

 

Safely? Yes (as long you understand that they are still not official drivers it is you who installs 3rd party drivers to your hardware.)
Would it be any use for you? No (because currently I am miles away still where we were before on modified driver builds)

 

Oh I missed you Maha Guru, and thank you for the new cleaner. Allways use your last version in between driver instalments and it's nice to have it updated.
Anyway, hurray you're back!

 

Allready installed your experimental release and first of all I'd like to say thanks to TWL for being back and I dont know if it's placebo or something, but that perfect mouse response I used to praise so much on your 5.10 and that after that was gone on every AMD´s release, IT'S BACK! Am I seing things? I dont know... Twl says he's done very little to this version (12.6a) but I had that exact driver installed just before this experimental one and the diference in smoothness and responsiveness is abismal. So far I've only tested Skyrim but this is what I have to say. My mouse is back to its prime in windows and in everything I do. THANK YOU right now.