|
|
| VDMuser |
| Posted: Jan 2 2013, 10:13 PM |
 |
|
Member
Group: Members
Posts: 24
Member No.: 25402
Joined: 11-April 09
|
Hello,
would it be possible to display the time shown in the job control windows for done jobs more detailed, i.e. with seconds or, maybe even better, an extra column that shows the duration of each job that is done?
Sometimes I test different codecs or different settings for the respective codec, and one thing that sometimes is crucial for my decision is the duration of the encoding process for different settings / codecs. Therefore it would be very helpful, if the job control window would show me the time in more detail (with seconds) or even directly show the duration for each finished job.
Thanks for any advice or help.
Mike |
|
| dloneranger |
| Posted: Jan 2 2013, 10:21 PM |
|
|
Moderator
Group: Moderators
Posts: 2366
Member No.: 22158
Joined: 26-September 07
|
Just as a general work around, if you right click on the finished file and select properties
You'll see 3 dates and times
Created is when the job started, modifed is when it was finished
(unless you've changed the files inbetween)
--------------------
MultiAdjust JoinWav WavNormalize FFMPeg Input Plugin v1827 UnSharpMask
Windows7/8 Codec Chooser
All FccHandlers Stuff inc. Installers for acm codecs AAC, AC3, LameMp3 |
|
| VDMuser |
| Posted: Jan 2 2013, 10:43 PM |
|
|
Member
Group: Members
Posts: 24
Member No.: 25402
Joined: 11-April 09
|
Ok, let me specify:
I'm only doing multi-pass encoding, so I would like to know how long exactly each pass took. That's an information that I can't get from the file itself. (Unless I would create a different file for each pass, which I clearly don't).
But when I save a job list with finished jobs there's a special code for the start and finish time stamp. I guess it would include seconds which are just not displayed in the job control window, but that's just my guess (that the time stamp code really includes seconds as information).
[Edit]
And as I'm at feature requesting anyway: 24-hour time display would be nice instead of "a" and "p" :-) |
|
| dloneranger |
| Posted: Jan 2 2013, 10:50 PM |
|
|
Moderator
Group: Moderators
Posts: 2366
Member No.: 22158
Joined: 26-September 07
|
Yeah - the start and end times are just hours and minutes
I'm not sure seconds would be all that useful, but can't see a reason not to have them it's a fairly trivial change
But, if your benchmarking comes down to comparing seconds than your error % gets pretty high
--------------------
MultiAdjust JoinWav WavNormalize FFMPeg Input Plugin v1827 UnSharpMask
Windows7/8 Codec Chooser
All FccHandlers Stuff inc. Installers for acm codecs AAC, AC3, LameMp3 |
|
| VDMuser |
| Posted: Jan 2 2013, 11:03 PM |
|
|
Member
Group: Members
Posts: 24
Member No.: 25402
Joined: 11-April 09
|
Well, the thing is, sometimes I use just short excerpt for testing and it makes quite a difference if the the start point says e.g. 6:00p and the end point e.g. 6:10p when in reality it might be 6:00:01 or 6:00:59 and 6:10:01 or 6:10:59, respectively. In one case it could be about 9 minutes of duration, in the other 11 minutes. And if the sample is just 10% or less of the whole video, extrapolating the encoding duration for the complete video (for comparision purposes with other settings) can be very misleading. Without the seconds displayed I can't know if the process started or ended closer to the earlier or later minute but due to rounding off maybe got set lower/higher at least half a minute or so. And with more than one pass that error could cumulate, making a good assessment nearly impossible.
I might "see" a difference in the duration of the encoding process when in realty there isn't any. Or the other way around.
That's the secret behind my "feature request"... :-) |
|
| dloneranger |
| Posted: Jan 2 2013, 11:36 PM |
|
|
Moderator
Group: Moderators
Posts: 2366
Member No.: 22158
Joined: 26-September 07
|
Oh, I understand the reasons 
To be honest, I really can't see the point though
10% is in the 'so small it hardly matters' range
Encoding usually falls into
Such a small clip it finishes almost straight away
So long I'm not going to sit and watch it
Or put another way
If by carefully testing I manage to get a better setting that is 10% quicker
Then something that took an hour now takes 54 minutes
or
For every 100 clips I encoded, I can now encode 110 clips
--------------------
MultiAdjust JoinWav WavNormalize FFMPeg Input Plugin v1827 UnSharpMask
Windows7/8 Codec Chooser
All FccHandlers Stuff inc. Installers for acm codecs AAC, AC3, LameMp3 |
|
| VDMuser |
| Posted: Jan 3 2013, 01:42 PM |
|
|
Member
Group: Members
Posts: 24
Member No.: 25402
Joined: 11-April 09
|
"10% is in the 'so small it hardly matters' range"
Well, then please wire me 10% of your money.
Would you consider 10% price inflation negligible?
Would you tell your boss "10% more money?
That's insignificant, don't bother"?
What if gas prices would rise 10%, would it hardly matter to you?
Have you ever gotten 10% interest on the money in your account?
Well, opinions are indeed quite diverse. I for one consider 10% huge.
Ask any businessman what he would think of 10% more sells, customers, returns... Or an IT specialist about 10% more computing power. Ask the people who invent mobile devices or batteries what they would give to have 10% more batterie running time.
Well, and me, I wouldn't mind to have 10% off my electricity bill or save me 10% time when encoding.
And to clear it up: I don't use VD for encoding or transcoding clips. I just use small chunks of longer videos for testing for quality, because testing the whole file with different setting might cost me a day or more. And at the end if I have to decide from the tests with the smaller chunks and have more than one candidate of codecs or settings I also want to know if the quality / size difference are worth a potential higher time consumption. Because time consumption in the end always equals money in one way or another. And I for one are NOT a millionaire who is able to not care about 10% more or less. :-) |
|
| dloneranger |
| Posted: Jan 3 2013, 08:04 PM |
|
|
Moderator
Group: Moderators
Posts: 2366
Member No.: 22158
Joined: 26-September 07
|
That's not quite what I was saying, but yup that about sums it up
Saving 2:25 hours on a 24 hour job is in my "shrug, was it worth the trouble in the first place" zone
And you've got to take away all the time spent testing and measuring, so that's even smaller
I'm having flashbacks to people disabling loads of windows services to get an extra 1fps in a game
Sorry if I sound really down on this as a concept, it's really hard to find a nice way to say it's something I wouldn't spend my time on
On the other hand, I know exactly what it's like to have someone look at me in complete bewilderment when I tell them what I've been doing with my time
Ah, that look of "uh, what? why would you do that????" on their faces - priceless :-)
I guess we're in one of those 'sports' moments where two people can't figure out why the other one likes the sport he does
You know the one
You're looking at them and thinking "seriously? you like volleyball"
And they're looking at you "what the hell? polo? those dudes on horses???"
Followed by that uncomfortable "........I have no idea what to say now" moment
Hopefully Phaeron will add the seconds for you :-)
--------------------
MultiAdjust JoinWav WavNormalize FFMPeg Input Plugin v1827 UnSharpMask
Windows7/8 Codec Chooser
All FccHandlers Stuff inc. Installers for acm codecs AAC, AC3, LameMp3 |
|
| VDMuser |
| Posted: Jan 5 2013, 03:38 PM |
|
|
Member
Group: Members
Posts: 24
Member No.: 25402
Joined: 11-April 09
|
You're right.
Well, the 10% were just an example. As were the up to 2 minutes (actually 1:58, presuming that the there is no rounding, just cutting off of the seconds of the time stamp) on 10 minutes converting time, which are actually 25%.
The point is, as long as there is no precise measurement, there's no way to tell about how much of an difference on is talking about.
But now for some slightly different approach:
I just had an idea. The "time stamp code" in the .jobs file shouldn't be a proprietary one, or is it? Maybe there's some translator for that language online were I just let the code be translated to human readable time stamps. That would help, too. |
|
| dloneranger |
| Posted: Jan 5 2013, 05:05 PM |
|
|
Moderator
Group: Moderators
Posts: 2366
Member No.: 22158
Joined: 26-September 07
|
It's the systems FILETIME
http://msdn.microsoft.com/en-gb/library/wi...v=vs.85%29.aspx
http://msdn.microsoft.com/en-gb/library/wi...v=vs.85%29.aspx
The first number is the high part and the second number is the low part (in hex)
| QUOTE | | A file time is a 64-bit value that represents the number of 100-nanosecond intervals that have elapsed since 12:00 A.M. January 1, 1601 Coordinated Universal Time (UTC). The system records file times when applications create, access, and write to files. |
So if you subtract the start time from the end time you'll have how many 100-nanosecond intervals the job took
IOW
(end time - start time)/10000000 = seconds
:-)
[edit] corrected the division
--------------------
MultiAdjust JoinWav WavNormalize FFMPeg Input Plugin v1827 UnSharpMask
Windows7/8 Codec Chooser
All FccHandlers Stuff inc. Installers for acm codecs AAC, AC3, LameMp3 |
|
| VDMuser |
| Posted: Jan 5 2013, 05:57 PM |
|
|
Member
Group: Members
Posts: 24
Member No.: 25402
Joined: 11-April 09
|
Thanks a lot for that info! I guess I'll try that. |
|
| Abrazo |
| Posted: Jan 5 2013, 08:12 PM |
|
|
Advanced Member
Group: Members
Posts: 775
Member No.: 28995
Joined: 5-November 10
|
If it still may be of some interest, this is a methode I found on the internet to show "seconds" when right-clicking a file in Windows 7 Explorer, and then clicking "Properties":
Basically, it involves changing the day of the week component of the "Long date" format found under Control Panel > Region and Language > Additional Settings... > Date tab.
By default, the day of the week component is represented by dddd. If you change it to ddd or remove it entirely, it should show you the exact time with seconds when you right click on a file and choose Properties.
So, like dloneranger said, if you calculate the time between "Created" and "Modified", that should result in the (more precise) time it took to create the file.
|
|
| dloneranger |
| Posted: Jan 5 2013, 08:21 PM |
|
|
Moderator
Group: Moderators
Posts: 2366
Member No.: 22158
Joined: 26-September 07
|
Not sure if this is any use to you
I just knocked up a little program that reads a jobs file and shows how long each finished job took
https://dl.dropbox.com/u/17024916/MyUploade...ff/JobTimes.zip
--------------------
MultiAdjust JoinWav WavNormalize FFMPeg Input Plugin v1827 UnSharpMask
Windows7/8 Codec Chooser
All FccHandlers Stuff inc. Installers for acm codecs AAC, AC3, LameMp3 |
|
| VDMuser |
| Posted: Jan 5 2013, 09:47 PM |
|
|
Member
Group: Members
Posts: 24
Member No.: 25402
Joined: 11-April 09
|
Well, dloneranger, if by "of any use" you mean "perfect", then you're right! :-)
Thanks a lot! Great little program! |
|
| VDMuser |
| Posted: Jan 5 2013, 09:54 PM |
|
|
Member
Group: Members
Posts: 24
Member No.: 25402
Joined: 11-April 09
|
Abrazo, sorry, but as we're talking about multi-pass encoding, file properties of the resulting video file won't help at all because they don't tell the story of the individual passes. It's all about the time stamps in the .jobs file. And dloneranger just solved the problem with his great program anyway. :-) |
|
