The 'Wish List' [merged]

[jdobbs]

Ok... I'll change that in the next version. Interestingly, if I added up all the changes that fall under the line "Other minor and cosmetic changes." in the release notes they probably add up to more than all the other changes combined. :)

[tom942]

I forgot other thing, don´t you mind to leave the final file created in the prepare phase when you are doing OPV or CQ to check quality, in the same way that RB-OPT?

I imagine that there are more cosmetic things in the TODO list than the important one's :).

[jdobbs]

I forgot other thing, don´t you mind to leave the final file created in the prepare phase when you are doing OPV or CQ to check quality, in the same way that RB-OPT?I don't think I understand.

[tom942]

Mmmm, maybe I express wrong :(.

When you select CCE and later "One pass VBR (w/analysis)" or the same for HC, after the prepare phase, RB put "Q value selected: 27" or other value in the log.

Well, I just saying not to delete that video sample, just to check the quality, or load it in bitrate viewer and see the quants, for instance :).

[tom942]

Using HC encoder in "One Pass VBR (w/analysis)", if you select "Menu encoding", the encoded menu is bigger than the original one and I suppose that it happens because it also encodes the menu in One Pass mode :(.

Could it be changed to multipass encode in the same way that CCE does?.

Thanks :).

[timbotu]

A suggestion to get rid of a lot of unnecessary clicks:

1) Have some kind of option to remove by default the "unreferenced audio" tracks on some dvds.

2) An option to automatically set the output directory to SpecifiedFolder\InputDirectoryName

3) Instead of "UNTITLED" make the default project file name that of the input directory.

That by itself would save a lot of clicks, especially when doing a bunch of movies at once.

[blutach]

1. The only way audio would be unreferenced is if ther stream can not be selected at all - I have never seen this in a commercial DVD and in any event, the remote could select the stream. If you are talking about audio not in your language, see Settings --> setup.

Regards

[timbotu]

I've seen both, but recent DVD-RB versions seem to have fixed it for one or the other. Maybe it's subtitles that need to be fixed. At any rate, it's something I noticed.

What I'd really like to see are the intuitive directory/saving ideas. I do 20 or 30 DVDs at a time and it gets really tedious.

[Boulder]

Stripped from another thread:
But I am working on ways to integrate external audio into the stream -- mainly as a path toward doing true authoring with the same engine. That's going to be a while, though.In addition to this, could you consider a feature to import external subtitles (or replace existing ones)? I have quite a lot of DVDs that have only hard-of-hearing subs which are quite annoying to use. It would help a lot if I had an automated way to import "cleaned up" subs.

[fjhdavid]

if it is not to difficult:

add a "correct audio delay" button using Nic (and jsoto) routines during the rebuild phase?

it should be helpful in order to synchronize audio and video

thanks
francois

[fjhdavid]

sorry for this second wish but I was thinking of an "automatic gain setting feature" for the movie audio channel, because when you record DVD, the dynamic range is not always pefect and dependant from the TV channel.

it could be add in an audio menu with the audio delay correction settings...

my two cents

but the software is wonderful as it is!

Francois

[Wombler]

Every so often I'm forced to leave the PC having already started 'One Click Mode' and would like to set the PC to shut down on completion.

I'd like to be able to alter the shut down options after 'One Click Mode' has been started but the menu with the 'Shutdown' option is locked and greyed out.

To make life easy would it be possible to have the 'On Completion' options made selectable during processing?

This would also cover the scenario where you'd set it to shut down and no longer wished to (I know there's already a box that appears at the end of processing but it would be nice not to have to wait for this to appear).

TIA.


Wombler

[RickA]

I would really like to see some kind of visual indicator that pc shutdown is activated. Nothing fancy mind you, just something that in a quick glance while the encoding process is underway to verify auto shutdown is active.

Usually, when I use this feature it is late at night before going to bed (brain is tired and groggy) or in a hurry out the door (too many things on the brain at once). Sometimes when I give my computer 'one last look' before turning off the monitor I wonder if I really did click on auto shutdown or not. Chalk it up to the onset of old age forgetfullness I guess. Still would be nice to 'know' that it is enabled without having to stop the process and restarting - just to make sure.

My apologies if already mentioned and my laziness for not reading through 19 pages to see if it has. :-)

Cheers,
Rick

[rack04]

:stupid: I would really use this feature.

[jdobbs]

For the next version the "Auto Shutdown" and "Close DVD-RB" functions have been moved to the FILE menu. That way you can drop the menu during execution and see immediately whether or not they are checked. You can also set or reset them as you wish.

[rack04]

For the next version the "Auto Shutdown" and "Close DVD-RB" functions have been moved to the FILE menu. That way you can drop the menu during execution and see immediately whether or not they are checked. You can also set or reset them as you wish.

Perfect. Thanks alot.

[Wombler]

For the next version the "Auto Shutdown" and "Close DVD-RB" functions have been moved to the FILE menu. That way you can drop the menu during execution and see immediately whether or not they are checked. You can also set or reset them as you wish.

Superb!

Many thanks for implementing this.:thanks:


Wombler

[rudiger]

Here's my entry on the wishlist: I would like an AppleTV option on DVDRebuilder Pro.

I just love DVD RB Pro! :thanks:

rudiger

[Rumbah]

It would be great if you could actually change "Auto Shutdown" and "Close DVD-RB" status while the encoding process is running.

That way you don't have to stop the encoding only because you decide to go to bed :) .

[Boulder]

Isn't that what jdobbs actually said: "That way you can drop the menu during execution and see immediately whether or not they are checked. You can also set or reset them as you wish."

[Rumbah]

Well, you're right, sorry about the previous post. I just didn't see the "You can also set or reset them as you wish".

[tom942]

Could you add a txt file in the same way than Redistribution.txt just for the classic method including the calculated bitrate of each segment?.

With that txt you could look quickly if you need to change the matrix for some segment, instead of search the bitrates through the ecl file.

[blutach]

@Tom

The easy way to see what the bitrates are for each segment is in the Preview/Edit tab. You can then decide if you wish to edit the ECL with a different matrix.

Regards

[tom942]

@blutach

Thank you for the tip. I don´t use to use the segment editor, and I didn´t knew that I could see it there :).

[Wishbringer]

Bitrate redist:
In prepare phase only one Instance of Encoder is used, even if multible cores are selected.
CPU usage is on Quadcore only 30% in that phase.

Encoding pipe on multicores:
On my quadcore HCenc is launced in encode phase four times in a row. And it counts through all segments in increasing segment numbers.
This could lead to suboptimal results in usage time when last segments are large. It last happened that only one Instance of HCenc was running such a large segment and all other Instances were finished.
Why not encode segments out of the row: largest first, then smaller ones?!

[blutach]

@Wishbringer

#1 is on the drawing board, I think as this has been mentioned recently.

#2 is a darned good idea, especially for episodic DVDs which tend to have long segments (to match their ad breaks). I'd like to second this idea!

Regards

[jdobbs]

I guess the driver for #2 would be "How often does that actually happen?" I can't remember it ever happening to me. So for me the bottom line is -- how much time do I want to spend writing something that (at least in my experience) almost never happens...

Anyone else experience this on a regular basis?

I'll look at it, though, and see if it is something that is easy to implement.

[blutach]

You do see it quite a bit jdobbs. I've seen some DVDs that only contain cells about 10 minutes long; particularly episodic DVDs or DVDs shot in inferior DVD cams or preprocessed by inferior package software that can only mark chapters at 5 minutes or so.

Now, you coud preprocess with VobBlanker or DVDRMP to split the last few cells down to "reasonable size" and this would take not much time (say 5 minutes), but this is probably as long as it takes to have the last encoder going by itself.

Regards

[jdobbs]

I guess I don't understand. I wouldn't have thought of 10 minutes as a long segment... and if they were all 10 minutes, then sorting them on size wouldn't make much difference.

[blutach]

Yes, if they were all 10 minutes, you're right.

But in the context of a regular feature movie, in my experience, 10 minutes is lengthy compared to most cell lengths (3-4 minutes) and in most episodic DVDs, there are lengthy segments but also little prologues (eg last week on name of show) = 4 of them times say 90 seconds each ... encode these last.

If the last segment is lengthy compared to at least one of the others, then there's only one encoder running over this last one, which takes a bit of additional time. It's not a big deal as it's only one segment, but the idea on a quad would be to have the machine working as hard as possible and not having 75% CPU resources free while encoding a long segment.

I suspect the time saved is only a few minutes overall, but this might be worth aiming for, if it is not too hard to implement (after all we know the segments' sizes before a frame is encoded).

Up to you, of course.

Regards

[blutach]

Just a note ....

HC022 now uses all CPU of multi processors, so this is now totally unnecessary.

Regards

[kumi]

a) I wish DVD-RB supported HC AQ.
b) I wish I could redistribute using HC "Fast" mode but encode using HC "Best" mode.
c) I also wish I could use other future obscure HC parameters at will :)

Maybe this could be accomplished with support for two override ini files:

C:\DVD-RB\HC1.ini
(used by DVD-RB in Phase I HC redistribution)
*AQ 0
*CQ_BFACTOR 1.4
*CQ_PFACTOR 2
*PROFILE FAST

C:\DVD-RB\HC2.ini
(used by DVD-RB in Phase II HC encoding)
*AQ 2
*CQ_BFACTOR 1.4
*CQ_PFACTOR 2
*PROFILE BEST

[Rippraff]

b) I wish I could redistribute using HC "Fast" mode but encode using HC "Best" mode.
July 7th, 2007 (v1.26.2)
CHANGES MADE TO PRO VERSION

- Turned off HC VBV and scene detection (except for a 100%
setting) during REDISTRIBUTION. This appears to have been
slowing down the speed of HC redist on some systems.
Also set HC to fast mode during the REDISTRIBUTION pass
as it gives the same results.

:)

Cu Rippraff

[kumi]

Doesn't work in one-click mode ;)

[lantern]

I have a suggestion for the wishlist. I know there wouldn't be too many people that would use it, but it would be nice to be able to select a custom size output from the DVD-RB interface instead of going into the ini file to change it. Maybe have it located with the DVD-5 DVD-9 selection as Custom Size and then it opens a dialog box to enter in sectors or MB.

Thanks

[tom942]

Hi,

I have noticed that when I do redistribution phase with HC (I use to use 100% sample instead than the default one), if LUMGAIN is ON, then that pass is very slowly and the new bitrate profile differs from CCE SP or SP2.

If I turn off LUMGAIN during redistribution pass it goes faster and the final bitrate profile it is very similar to the one obtained with CCE SP or SP2.

Could you add something that:

1) Turns off LUMGAIN during redistribution pass (at least with a 100% sample)

2) and then turns it on during encode phase, with the value selected in the DVD-RB?.

Thanks and greets.

[Sharc]

I have noticed that when I do redistribution phase with HC (I use to use 100% sample instead than the default one), if LUMGAIN is ON, then that pass is very slowly and the new bitrate profile differs from CCE SP or SP2.

If I turn off LUMGAIN during redistribution pass it goes faster and the final bitrate profile it is very similar to the one obtained with CCE SP or SP2.

You may add in the rebuilder.ini under [Options]:
REDIST_USE_FINAL_Q=1
(It normally makes the profiles of HC and CCE look more similar)

Could you add something that:

1) Turns off LUMGAIN during redistribution pass (at least with a 100% sample)

2) and then turns it on during encode phase, with the value selected in the DVD-RB?.

Hmmm... , not sure if this is recommended in general, because for best results the redistribution should be run with same (or most similar) conditions as the final encode for best results. Otherwise there is a mismatch between the two.

[Boulder]

LUMGAIN does nothing in CQ mode, it only affects the 2-pass encode.

[Sharc]

So does this mean in can be disabled for the redistribution pass without consequence on the redistributed bitrate per segment for the final encode (other than saving time?).

[Boulder]

Yes, it doesn't matter whether it's enabled or not.. You can check the behaviour by enabling LUMGAIN and then opening the resulting m2v file in DGIndex and using the Log Quant Matrices option while saving a project. I once tested this and the quant matrix didn't change at all in the CQ encode.

[Sharc]

Thanks, good to know.
(Now I am just wondering a bit why LUMGAIN takes extra time when enabled for the redistribution pass, according to tom942's observation. Maybe the algoritm is still running but eventually has no effect on the matrix?).

[Boulder]

Hey..I stand corrected :) It seems that the feature was disabled for CQ mode in some earlier version but in the latest 024 beta, it does work. Which is actually really nice, I think I posted about this issue in some HC thread long time ago.

[tom942]

Hi Sharc,

I've been a long time user of that option :). I still remember the discussion between if it was worth or not the time spent in the previous passes looking for the FINAL_Q, or just use BASE_Q :).

I thought as you, that it was better to have in both phases LUMGAIN ON, but the time for the redistribution pass was much more longer that with it OFF.

Then I did several different tests and later compare the Resditribution.txt from HC with LUMGAIN ON, later OFF and to the ones generated using SP or SP2.
If LUMGAIN was ON, the difference with the profile with the one generated by CCE family in the same cells sometimes can go up to about 500 kpbs.
With it OFF, it is more close to the one generated by CCE.

Then the reasoning that I followed was that if CCE didn´t modify its matrix during redist phase, and later, in the encoding phase, you could use AQM without harm, then I could do something similar with HC (Phase one, LUMGAIN OFF, Phase two LUMGAIN ON).
Also, as we can use HC to avoid CCE "bug" and later encode with CCE own settings, then I thought that we could use/translate similar settings.
For instance, I remember that the GOP size in redist phase is fixed to 12, although you've got for encoding phase, 15 or auto.

I can do this manually but I prefer that DVD-RB does it automatically, mainly if I forget to change the setting between a phase and the other.

[tom942]

Hi Boulder,

I don´t know what version you tested, but 0.23 (I don´t know about 0.24 because I don´t have tested yet, what a shame :p) if you look at the GUI while is in CQ mode, you'll see in the matrix section the "ADAPTIVE" word.
Later, as you said, if you check the log generated by DGIndex, you'll see that there are several matrices, not just the one selected.