Feature Requests
- AYColumbia
- Posts: 88
- Joined: Sun Feb 12, 2012 11:29 am
Feature Requests
Thanks for the response. No worries, was just curious and figured it may be a lot of work. I rarely see VP9 used honestly.
Feature Requests
Yeah, it's typically used for streaming by some providers. AV1 is also something to consider supporting. Who knows, one day I may get a burst of energy...or a contract. ![Wink ;)](./images/smilies/icon_e_wink.gif)
![Wink ;)](./images/smilies/icon_e_wink.gif)
Feature Requests
Would it be possible to index a Dolby Vision FEL substream at the same time as its HEVC parent stream? So we can have 2 DGI index files that link back to the same source m2ts file(s), for easier use with something like DoviBaker
Currently I'm demuxing the Enhancement layer for each disc, per DoviBaker's instructions, and then indexing/referencing that. But it's just eating up extra disk space, when all the data it needs is already in the m2ts. Some of these tracks get up to 8-10GB!
Currently I'm demuxing the Enhancement layer for each disc, per DoviBaker's instructions, and then indexing/referencing that. But it's just eating up extra disk space, when all the data it needs is already in the m2ts. Some of these tracks get up to 8-10GB!
Feature Requests
Hello PoeBear. Nice to hear from you again.
Open the disk. Index the BL as usual. Then go into Stream/Set PIDs and change the Video PID to that of the EL stream. Now index again, using a different name for the DGI.
That will give you two DGIs referencing the original M2TS(s), one for the BL and one for the EL. That saves the disk space as you requested. Sorry but I cannot do the two indexing operations simultaneously.
Open the disk. Index the BL as usual. Then go into Stream/Set PIDs and change the Video PID to that of the EL stream. Now index again, using a different name for the DGI.
That will give you two DGIs referencing the original M2TS(s), one for the BL and one for the EL. That saves the disk space as you requested. Sorry but I cannot do the two indexing operations simultaneously.
Feature Requests
Cheers! This worked, and outputs exactly the same in my frameserver
![grateful :salute:](./images/smilies/icon_salute.gif)
Would it be possible to have the Dolby Vision RPU demuxed as part of the indexing/demuxing process, or is how its embedded into the EL necessitate demuxing first?
Feature Requests
Please use DDVT or dovi_tool for that. I don't have the knowledge or energy to try to implement that. ![Sad :(](./images/smilies/icon_e_sad.gif)
![Sad :(](./images/smilies/icon_e_sad.gif)
Feature Requests
For streams with multiple videos, it is currently not possible to demux any but the first video from the stream. Would it be possible to allow for demuxing also the others?
Feature Requests
It's a similar question to that discussed above with the DolbyVision BL and EL.
Go into Stream/Set PIDs and change the Video PID to that of the desired stream. You can't do multiple streams simultaneously, however.
Go into Stream/Set PIDs and change the Video PID to that of the desired stream. You can't do multiple streams simultaneously, however.
Feature Requests
Is it possible to boost DGSharpen() a bit more in 246? I like the new one but it's definitely more subtle at 1.0 than the old version & seems like the threshold for light maybe lower...or higher, i always forget which way thresh works...it seems not as intense rn.
Feature Requests
Thank you for your feedback, Whackbag. I will look into it. Possibly we can offer different kernels controlled by an option.
Feature Requests
Something good coming your way, Whackbag. Thank you for your patience.
Feature Requests
Here is a test version for DGDecNV including an improved sharpener DGSharpen(). It also includes the DTS identification fix for DGIndexNV. The new sharpener has a more useful strength function to get stronger sharpening, overshoot/undershoot parameters to allow exceeding the local limit, and edge mask options to exclude ranges from sharpening (e.g., to avoid artifacts at the edges of letterboxing). Please refer to the included manual for details on the new parameters.
Your feedback will be appreciated. I will release this as build 247 after it is verified OK.
http://rationalqm.us/misc/DGDecNV_test.zip
Your feedback will be appreciated. I will release this as build 247 after it is verified OK.
http://rationalqm.us/misc/DGDecNV_test.zip
Feature Requests
Cheers Rocky, i'll play around with it for a couple days and see how she fairs...thanx again.
Feature Requests
Thank you, Whackbag. BTW, if you set strength = 5 and everything else 0 it should be the same as the original (not the last iteration) with strength 1.0. Now you can go sharper if you wish, and you can violate the limiting in a controlled fashion.
Feature Requests
245 (strength=1.0)
![Image](https://i.ibb.co/bFw39Zh/245-1-0.png)
246test (strength=13.0, os=7)
![Image](https://i.ibb.co/fMwzRZp/246-strength-13-0-os-7.png)
I like the over sharpened look & i use a cube file not as intended, i don't use it for color correction but as a slight denoiser. So this is not a true apples=apples comparison as i figured if i was going to rerun a bunch of tests i may as well tweak some other stuff while i was at it (slight color change is from a chroma-qp-offset i changed in Staxrip).
So yeah...i'll keep playing with it for a bit...but i think this works for me.
![Image](https://i.ibb.co/bFw39Zh/245-1-0.png)
246test (strength=13.0, os=7)
![Image](https://i.ibb.co/fMwzRZp/246-strength-13-0-os-7.png)
I like the over sharpened look & i use a cube file not as intended, i don't use it for color correction but as a slight denoiser. So this is not a true apples=apples comparison as i figured if i was going to rerun a bunch of tests i may as well tweak some other stuff while i was at it (slight color change is from a chroma-qp-offset i changed in Staxrip).
So yeah...i'll keep playing with it for a bit...but i think this works for me.
Feature Requests
Super. Thank you for your testing.
Feature Requests
I did had again a user who enabled DGDecNV without the proper hardware. It would be great if there would be a possibility to implement some kind of check if DGDecNV is suported on a system. If that check succeeds I could enable DGDecNV automatically and if not, it cannot be enabled/used.
E.g. DGIndexNV.exe will be called with "-check" and will return a string with the supported codecs (so none if card too old or no Nvidia) and ExitCode 0 if supported and 1 if not supported at all?
This is a "nice to have" feature and for sure there are more important things than this here![Thanks :hat:](./images/smilies/icon_hattip.gif)
E.g. DGIndexNV.exe will be called with "-check" and will return a string with the supported codecs (so none if card too old or no Nvidia) and ExitCode 0 if supported and 1 if not supported at all?
This is a "nice to have" feature and for sure there are more important things than this here
![Thanks :hat:](./images/smilies/icon_hattip.gif)
Feature Requests
Hi Zathor. Will you actually use the codec info? It will be much easier for me to implement a simple yes/no.
Feature Requests
In my current implementation: no.
Potentially I would have implemeted something so that the filter is enabled for specific codecs depending on if it is supported or not. However that would also cause some effort on my side with likely low benefits so if it is causing more effort also on your side then for sure the information if the system is supported or not would be more than sufficient.
Potentially I would have implemeted something so that the filter is enabled for specific codecs depending on if it is supported or not. However that would also cause some effort on my side with likely low benefits so if it is causing more effort also on your side then for sure the information if the system is supported or not would be more than sufficient.
Feature Requests
OK, thank you. It's hard to "return a string" but I can do a series of return codes:
0: DGDecNV not supported
1: MPEG2
2: MPEG2/AVC
3: MPEG2/AVC/VC1
4: MPEG2/AVC/VC1/HEVC
0: DGDecNV not supported
1: MPEG2
2: MPEG2/AVC
3: MPEG2/AVC/VC1
4: MPEG2/AVC/VC1/HEVC
Feature Requests
That sounds great and would be very nice. So even the supported codecs are possible to detect? Even better ![Thanks :hat:](./images/smilies/icon_hattip.gif)
![Thanks :hat:](./images/smilies/icon_hattip.gif)
Feature Requests
The initial implementation will give a yes/no indication only.
When this option is given DGIndexNV checks whether its operation is supported on the hardware (it tries to initialize CUDA). DGIndexNV returns immediately with exit status 1 if operation is supported and exit status 0 if operation is not supported. Because DGIndexNV is a windows application the error status must be checked differently. Do this:
start /wait DGIndexNV.exe -check
echo %errorlevel%
Released in build 248. I hope it is usable for you. If not, please let me know.
When this option is given DGIndexNV checks whether its operation is supported on the hardware (it tries to initialize CUDA). DGIndexNV returns immediately with exit status 1 if operation is supported and exit status 0 if operation is not supported. Because DGIndexNV is a windows application the error status must be checked differently. Do this:
start /wait DGIndexNV.exe -check
echo %errorlevel%
Released in build 248. I hope it is usable for you. If not, please let me know.
Feature Requests
Thank you very much
I just tested it on my system with a GTX 1060 ad driver 531.79 and the exit code is 1. Based on the list provided by you above I would have expected a 4 or am I wrong? It is not a big deal as currently I would just detected for 0 or > 0.
Additionally I tested it on a system without Nvidia card and the result was as expected a 0. But two error windows showed up telling me that nvcuvid.dll and nvcuda.dll cannot be found. Would it be possible to hide those?
![Thanks :hat:](./images/smilies/icon_hattip.gif)
I just tested it on my system with a GTX 1060 ad driver 531.79 and the exit code is 1. Based on the list provided by you above I would have expected a 4 or am I wrong? It is not a big deal as currently I would just detected for 0 or > 0.
Additionally I tested it on a system without Nvidia card and the result was as expected a 0. But two error windows showed up telling me that nvcuvid.dll and nvcuda.dll cannot be found. Would it be possible to hide those?
Feature Requests
As I said above, the initial implementation will give a yes/no (0/1) indication only.
I'll see if I can stop the DLL errors. Probably gonna have to delay load them, etc.
I'll see if I can stop the DLL errors. Probably gonna have to delay load them, etc.
Feature Requests
Please re-download 248 and test. I tested on our main machine (returned 1 as nVidia is present), and on a laptop (returned 0 as nVidia is not present).