Colorizing black & white movie

[Roméo Latour]

Hi,
I would like to colorize one of my w&b movies.
I'm looking for a software dedicated to doing that job.

However, I do not want to colorize the fames one by one.
I'm looking for a software that does the job shot by shot.
This is how the pros do it these days, but not sure this is available to the general public.
Thanks.

[johns0]

Unless you want to spend a ton of money there's no software that will do what you want.

[redwudz]

Some background: https://forum.videohelp.com/threads/268985-Film-colorization-how-made

[pandy]

Hi,
I would like to colorize one of my w&b movies

This is how the pros do it these days, but not sure this is available to the general public.

It is very easy - outsource this to specialized company in India - they use humans as neural network machine - pros call this 'resource' or 'asset'.

[themaster1]

i remember that someone had developped a software (or algorithm) to recover colours from black & white movies unfortunately the page has disapeared it seems, it was here: http://colourrecovery.wikispaces.com/?responseToken=bb0997c14313663f2c0c5ad547c60f99

There is an article talking about it here: http://www.theguardian.com/technology/2008/dec/11/digital-video-restoration-dad-s-army

[jagabo]

i remember that someone had developped a software (or algorithm) to recover colours from black & white movies

That only works for B&W film that was made from color video tape. Some of the video chroma subcarrier is still visible on the film. It does not work on material that was shot on B&W film.

[lordsmurf]

unfortunately the page has disapeared it seems, it was here: http://colourrecovery.wikispaces.com/?responseToken=bb0997c14313663f2c0c5ad547c60f99

Try here: https://web.archive.org/web/20120929012232/http://colourrecovery.wikispaces.com/

Good ol' U.S. LOC Archive.org Wayback!

[pandy]

Same as jagabo.
I believe that OP express interest to colorize BW movie not recover color information coded in BW video during process transferring from video magnetic tape to cinema tape (yeah - ages ago magnetic tape was expensive and common practice was to transfer recorded content to cheaper cinema tape and as machine use to this was monochromatic then chroma subcarrier coded color information was transferred also so recovery is possible).

[bmick23]

You want to use Mocha Pro, a motion tracking software. This will allow you to track each object and create a mask that you can import into After Effects to add color. I use this exact process and it works very well.

Here is my demo reel of colorized footage using this method:
https://www.youtube.com/watch?v=ZPJMZqK6ZMY

[manono]

Here is my demo reel of colorized footage using this method:
https://www.youtube.com/watch?v=ZPJMZqK6ZMY

It looks like crap, if you ask me, but I'm not a fan of colorization in any way, shape or form. You should try and learn how to IVTC a film source rather than use a poor deinterlacer, before attempting 'advanced' video editing. It's supposed to be 23.976fps and not 29.97fps. There's a ton of aliasing and every fifth frame is a duplicate, making the whole thing play 'stuttery'.

[bmick23]

Here is my demo reel of colorized footage using this method:
https://www.youtube.com/watch?v=ZPJMZqK6ZMY

It looks like crap, if you ask me, but I'm not a fan of colorization in any way, shape or form. You should try and learn how to IVTC a film source rather than use a poor deinterlacer, before attempting 'advanced' video editing. It's supposed to be 23.976fps and not 29.97fps. There's a ton of aliasing and every fifth frame is a duplicate, making the whole thing play 'stuttery'.

The demo is for the colorization, not frame rates.

[jagabo]

Here is my demo reel of colorized footage using this method:
https://www.youtube.com/watch?v=ZPJMZqK6ZMY

Just curios... How long did it take to colorize that short clip?

[bmick23]

Here is my demo reel of colorized footage using this method:
https://www.youtube.com/watch?v=ZPJMZqK6ZMY

Just curios... How long did it take to colorize that short clip?

I did them all at different times, but it depends on the complexity and how much movement there is in the shot. The full clip of the awards speech at the end is 30 seconds long and it took about 4/5 hours. Lynda.com has some great courses to learn the Mocha software.

[jagabo]

Here is my demo reel of colorized footage using this method:
https://www.youtube.com/watch?v=ZPJMZqK6ZMY

Just curios... How long did it take to colorize that short clip?

...The full clip of the awards speech at the end is 30 seconds long and it took about 4/5 hours.

Thanks. Though I timed that shot as ~15 seconds.

[bmick23]

Here is my demo reel of colorized footage using this method:
https://www.youtube.com/watch?v=ZPJMZqK6ZMY

Just curios... How long did it take to colorize that short clip?

...The full clip of the awards speech at the end is 30 seconds long and it took about 4/5 hours.

Thanks. Though I timed that shot as ~15 seconds.

The full clip is 30 seconds. I only used half of it in the demo.

[bmick23]

Unless you want to spend a ton of money there's no software that will do what you want.

Everything OP wants to do can be done using After Effects and the bundled Mocha plugin. A license is $19/mo. Not that expensive. Why not be helpful instead of dismissive?

[johns0]

Because we have had a lot of posts asking this question and it was always the same answer,if you found a cheaper way to do it great,still gonna take a lot of time and money since isn't labor considered money?

[DB83]

These later posts smell like a commercial.

One thing about proper colorization is skin tones. These are practically non-existant in these samples.

And the OP wants to do this automatically. Not spend hours over a few seconds. I could do that in my old NLE.

Never been a great fan of the process - made many posts about that here before. Maybe my opinion was poisoned by some early botches. Watched a recent one of 'Scrooge' - the Alistair Sim - one and that is watchable. But I would still watch the original B&W on the same dvd given the choice.

[johns0]

Ya,the Marilyn Munster shots looks unnatural like she has a flat face,as i said before to do a proper job takes much better software/equipment and time.

[bmick23]

Ya,the Marilyn Munster shots looks unnatural like she has a flat face,as i said before to do a proper job takes much better software/equipment and time.

All it takes is a decent setup, a color-corrected monitor and the right software. Colorization isn't complicated, it's just time consuming because there are so many elements. However you want to add color (After Effects, Resolve, whatever), it can be done well and within a reasonable budget. All of the demo footage I posted was done while I was learning the software, so of course it's not the best. Skin tones can be done well and fairly easily using color curves or even multiple masks layers. The biggest cost is time, but if you're learning for fun like me, it's worth it.

[-Habanero-]

There are other methods but you'll still have to mark every 6th frame or so in a movie for best results.
http://www.cs.huji.ac.il/~yweiss/Colorization/index.html#video

[Roméo Latour]

There are other methods but you'll still have to mark every 6th frame or so in a movie for best results.
http://www.cs.huji.ac.il/~yweiss/Colorization/index.html#video

This sounds like what I'm looking for.Is there a software and tutor that could go with.
Thanks to all those of you who fed this topic and sorry for being late.

[-Habanero-]

You need MATLAB to use it. I've done it only once a long time ago for a quick touch-up job so I forget the specifics.

[johnmeyer]

Much to my utter amazement, there may actually be an answer to the OP's original request. Read on.

I knew about the BBC efforts to recover color using residual timing information in old Kinescopes.

However, I just read a post over in the doom9.org forum where someone has demonstrated an algorithm, using only a few dozen lines of AVISynth code, that appears to be able to recover color from B&W, without any such residual timing information.

Here's the link:

automated, color, colorization Automated Colorization Script

The second person in this thread is someone I have worked with on several projects. He is a great programmer, and utterly reliable. He has been able to duplicate the OP's results. The colors in the resulting video are far from perfect, and won't compare to what you would get with traditional labor-intensive colorizing efforts already described earlier in this thread. However, it is pretty much exactly what the OP asked for.

The script used is short, but very obscure. So far no one has been able to explain how it works.

[Cornucopia]

As someone who's worked at colorizing projects manually for well over a decade, I would say this is interesting, but far from amazing. And unfortunately ultimately crappy and unusable except for 10% of material - more than half of which it would mainly serve as a first pass prior to the greater manual work. (Look at all the chroma smearing!!)
The algorithm is making uneducated & premature assumptions about the relationship between certain luma levels and some associated chroma designation, but there are plenty of possible examples of existing work whose luma levels should suggest a wholly different chroma result.
I think this says more about the choice of test materials and the mindset of those making the assumptions than it does to the glory of the algorithm.
Similar things have been attempted with automatic 3D creation with similar bad psuedo results.

IOW, nice try but that dog can't hunt.

Scott

[poisondeathray]

I agree with Scott 100%

I don't want to sound like a downer - while the avisynth method is cool for type of approach, it's really not that impressive - it fails miserably in many shots. The accuracy and masks can be quite innacurate, and there appears to be no easy way to tweak it or fix it - it's true of avisynth in general. I suppose that's expected for this type of approach, but avisynth is still very useful for other manipulations

The OP was asking for scene to scene, I doubt this is what he was looking for

[johnmeyer]

Well, I thought it was worth bringing to his attention because everyone, including me, was saying that it was impossible to do anything automatically. I was already aware of the limitations you both mentioned, and said so in my post above. However, the fact that it can recover any color at all, and that it can get many of the colors correct is pretty amazing.

If nothing else, it would provide a great starting point for doing the "traditional" hand-colorizing because it seems to do a pretty good job getting the correct hue for most of the objects in the test clips. For example, the carrots are orange, and box in the background is blue, etc. It would be a lot easier to do a colorizing job if objects were already close the right hue, and if the color tracked the objects reasonably well, as it appears they do.

[poisondeathray]

If nothing else, it would provide a great starting point for doing the "traditional" hand-colorizing because it seems to do a pretty good job getting the correct hue for most of the objects in the test clips. For example, the carrots are orange, and box in the background is blue, etc. It would be a lot easier to do a colorizing job if objects were already close the right hue, and if the color tracked the objects reasonably well, as it appears they do.

Really, I don't think it helps any in a "traditional" coloring workflow. Most people already know carrots are orange etc..

I'm banging my head trying to think of how this be incorporated in a production workflow, even if it could shave off 2% time in selected shots it could be useful - but it can't - the reason is the raw mask data won't be accessible in the other host program.

The reality is it would take more time fixing things, because the "off" line boundaries can interfere with real trackers. It's like some kindergarten kids colored over the lines. The original B/W is presumed accurate, but now it has just just smudged everything and made it more difficult

Still very cool, and I'm keeping an eye on it to see what develops

[Cornucopia]

But let's get it very clear & agreed upon here (for all the newbies checking into these threads): there is NO "recovery" of colors with this method, or any method given a true monochrome source. It's just guesswork "synthesis" (or re-synthesis, if you like).

Scott

[johnmeyer]

Wow, I sure am glad I posted.

I'm am completely surprised by both the tone and the negativism of the replies posted above. I am especially puzzled because even traditional B&W colorization is pretty awful and, IMHO, a complete waste of time. In fact, I avoid all colorized movies because they look so bad. To criticize this method by saying that it sometimes chooses the wrong color is completely ludicrous because the "real" colorization method gets it wrong all the time. What's more, while the colors produced by this method seem to have a purplish cast, and don't perfectly track the object, that has been my experience with colorized movies as well.

The differences are a matter of degree.

So, for the third time, for those who apparently didn't read my two previous posts or simply chose to ignore them for some reason, I am making no claim that this is some sort of ultimate colorization technology, and it has tremendous limitations and shortcomings . Unfortunately I have a sneaking suspicion that the same people will once again ignore this statement, even though I have shouted it this time by putting it in boldface.

Finally, to get back to the main point, this is almost certainly the only method available to some guy (the OP) who wants to colorize his own home movies. What else can he do on a zero budget, and with little time to spend? Certainly no one else in this thread , including those who just posted above, have suggested any other technique that he is likely to be able to use, including the two who seem to have gotten upset, for reasons that are completely unclear to me.

So, calm down, and let the OP take a look and decide for himself. If he too thinks the technology is useless, let him make that decision. I actually don't understand the point of your "warnings" for newbies. Are you saying that they'll try this and think it is great? If they do actually try it, and if they like the results, isn't that a good thing? To say otherwise -- that you know better than they do about what they should like -- is the definition of snobbery.

I'm just trying to help the guy, so please, give us all a break!!