Messages posted by Elochai

Quote

no solution yet till I can get it to compile. Yeah I thought the f was weird too. I narrowed it down to the 1st char in the command line after the -i for the input file which is frames.txt. If I change it to a full path, it’ll be the drive later inside of the f.

not sure why it be the padding now though as I eliminated that completely by inputting full file names for all frames. So it doesn’t need the %d at all. It’s reading the full path and file on each line of the frames.txt file.

Got it figured out.

It is the 16-bit TIFF files that are the issue. They're not being encoded correctly for ffmpeg to work with them.

Quote

No, don't think tiff is the issue. I can esseentially mimic your initial error

[image2 @ 000001f43c953380] Could find no file with path 'C:\Users\Jeff\ffmpeg\one\img_%d.jpg' and index in the range 0-4
[concat @ 000001f43c94a340] Impossible to open 'C:\Users\Jeff\ffmpeg\one\img_%d.jpg'
files.txt: No such file or directory

That has to do with padding as I mentioned and index.

Any chance you simply have a hack in your file? Why is location reference 'fJ:\video\' in error msg?

Anyhow, glad you have a solution. I really doubt you'd be anywhere near solution with PD and 80,000 pics in the timeline. I had hopes they would improve timeline response for such tasks when they added the inverse, produce image sequence at PD17, but they didn't.

Jeff

no solution yet till I can get it to compile. Yeah I thought the f was weird too. I narrowed it down to the 1st char in the command line after the -i for the input file which is frames.txt. If I change it to a full path, it’ll be the drive later inside of the f.

not sure why it be the padding now though as I eliminated that completely by inputting full file names for all frames. So it doesn’t need the %d at all. It’s reading the full path and file on each line of the frames.txt file.

Quote

Mine appears to work fine with padding.
files.txt
file 'C:\Users\Jeff\ffmpeg\one\img_'%03d'.jpg'
file 'C:\Users\Jeff\ffmpeg\two\img_'%03d'.jpg'

Since start not specified second group of files needed to be from 000 again, I had just 200 pics in one and 382 pics in two.

Jeff

yeah, didn’t work for me, but I got past that by doing each frame by file name.

but as you can see, it see’s the files but won’t work with them. Starting to wonder if it has to do with the tiff file format or how they got compiled.

Quote error looks like index issue, 0 padding on %d, %06d

Jeff

Yeah I thought that too and changed it to %06d but no luck, same error. Also tried 0%d as the 1st 0 doesn't change in the file names. That also didn't work. Same error.

I have now generated a file with every frame in it so I don't have to use the %d command.

Now I get this error.

[concat @ 000001d30cddc840] Impossible to open 'fJ:\video\batchone\000000.tiff'
frames.txt: Protocol not found
Did you mean file:frames.txt?

Quote

There is a limit of 2500 files not 25000 in the media library. You never supplied any relevant information or needed details. LOL.

Sorry I misread the error by one 0 ... What relevant information where you looking for ? Upon till that point, I didn't know PD had a limit to the number of files it could take into a project so I had no need to specify that I have a project with 79,820 frames that spans across two hard drives and takes up a total of 2.74 TB in 4K frames. Kinda figured the limit would of been my hardware ability. If you mean version of PD, it 16 and yes it has H.265, that what I use to encode 99% of my video as it produces a better file size.

Quote
The link showed you how to get around the 2,500 limit in the media library.

Regardless, with 80,000, as I had suggested, my experience is PD is
not for that task. I typically do about 10,000 in VirtualDub, I see no
reason it couldn't handle more though, ffmpeg should be similar
experience.

Jeff

Yeah no, I think your right about PD not being the best choice for doing the frames, I didn't even bother with the link after I tried importing the frames into PD, it gonna be slugish and I don't think it will result well.

Right now I am trying FFmpeg, even if I can get the footage done, then I can add the sound in after. But I am running into an error and can't seem to figure out where my mistake is. May have to check over there to see what silly thing I am doing wrong. 1st time using FFmpeg.

Got a concat file called frames.txt that has my frames locations

file 'J:\video\batchone\%d.tiff'
file 'F:\video\batchtwo\%d.tiff'

So from my understadning this should allow it to go from the first frame (000000.tiff) to the last frame (079820.tiff)

Then I am using the comand:

ffmpeg -r 30 -f concat -safe 0 -i frames.txt -vcodec libx265 -crf 20 -pix_fmt yuv420p F:\VideoOutput.mp4

but I then get an error

[image2 @ 00000293e9826c80] Could find no file with path 'fJ:\video\batchone\%d.tiff' and index in the range 0-4
[concat @ 00000293e981c940] Impossible to open 'fJ:\video\batchone\%d.tiff'
frames.txt: No such file or directory

After checking the file names, and paths, I'm sure it something simple I am over looking.

Quote

ffmpeg, VirtualDub, probably anything better than PD for that task.

It is possible, I showed that yrs ago here: https://forum.cyberlink.com/forum/posts/list/31108.page#post_box_200199 while others doubted. It took considerable patience then as Tony replicated and discovered, and not much better with a more recent PD version today. PD program response with hundreds, or thousands of pics in the timeline is not productive at all. So it really depends on how many tiff files, I often use VirtualDub for the task with several thousand photos.

You can use H.264, H.265 or any other format in PD if you wish, the output when carefully inspected may not be what you desire though, with any default profile, it won't necessary be the exact timeline frames in that order put to video even if each tiff file is one frame.

Jeff

Thinking I'm gonna have to try this with ffmpeg, never done it before but PD gives an error about being over it max supported file number of 25,000 files when importing the frames. I have just shy of 80,000 frames.

So not sure if this can be done in PD or not. But I have a full project of a video output by each frame to .tiff format after working on each frame.

So I want to import all frames into PD along with the audio file that goes with it, and then produce the video back into a playable format using H.265 (Not H264) codec. Can this be done by PD, or an I better off looking into ffmeg for this ?

Quote

Elochai, I'd assume probably Topaz Labs. If so, you can still get their free beta version (google search) which is probably not nearly as refined of a release but I've had good luck with it. Has a small UI but I predominately use in cli so not for everyone. I have found it better than the other product you mention for most stuff I've done.

Since a competitor with PD's "Video Enhancement" technology which is a very poor upsampling and sharpening technology one probably can't write too much or the thread will be locked and/or removed.

Jeff

You would be correct, I am using Topaz Labs Enhance AI and it blows DVDfab version out of the water.

Can't really post any video clips of my convertions as people on the videos wouldn't like that to much. But nothing stopping people to try it out.

I can't see CyberLink locking or closing the thread tho as they don't have a product that does the same thing. PD's "Video Enhancement" technology doesn't use AI or machine learning to generate new upscaled video.

This is a great tool tho to upscale your low res videos and then work on them in PD.

I was gonna post a few frames from an old 1993 clip homevideo but it looks like I would have to host them myself or go through the trouble of an image host for these forums.

So far I only used the Gaia HQ and Gaia-CG for the AI option, the other two Artemis options for AI seem to result in a lot of noise and blocks during upscale.

I have found Gaia-CG (The version of AI that computer generates new pixels) to work very well when it comes to faces in a video being upscaled. I have footage where even the 480p version of the footage has a lot of blocky-ness on a face and it upscaled it and generated the face to near prefect. Was better then the original. Now yes I also got results where you can tell it had pixels added in but the results were still better then no upscaling at all.

I was more amazed yesterday (as I said I was upscaling a 1 hour and 30 min video for 14 hours, well it was done after 20 hours total upscale time). In one of the videos, there was a wedding dress with lace. On a close up shot, the lace was very burry and blended into a blocky pixelated area. When upscaled with Gaia-CG option, it came out with each area of the lace highy visible and detailed. each hole and lace pattern clear and shape.

Quote

since panny upscales VHS, DvD and SDHC to 720p 1080p 1080i to the HDMI output... maybe you can put in something in between to capture upscaled video like Blu-ray recoder???

i guess, it comes down to how bad you want it. ^^

happy happy joy joy

PepsiMan
'garbage in garbage out'

yes, like I was saying, I don’t need a physical upscaler as I have that now. It software I need to upscale the 480p files I have in MP4 format on the computer already.

Good news is, I found software that is capable of doing it using AI technology. So far the results are 120% better then what DVDfab has to offer for there AI upscaler. Bad news is that it costs $200 USD or for me about $282 Canadian.

But the results are extremely good, intense on hardware tho, I’m still upscaling a 480p file to 1080p that is over 1 hour and a half of footage, about 14 hours into the AI upscaling but like I said, the results look very good, looks as if it was shot with a 1080p camcorder and the AI is doing the pixels (adding in and smoothing our areas) that even hard to see in the original footage.

Quote i have panny dmr-ez475v that'll upscale 480p to 720p 1080p 1080i through the HDMI...
go to panny's Product Guide - DVD Recorder DMR-EZ475 to see what you can tinker with and buy a used one in decent condition on eBay.
yup, mine is about 13 years old and still kicking. i haven't replaced any pinch rollers yet... plays the good, the bad and the ugly in VHS tape... it'll record VHS-DvD and SDHC-DvD. ^^

happy happy joy joy

PepsiMan
'garbage in garbage out'

yeah, I got a upscaler for that by WavLink. What I’m looking for is software to upscale 480p video that I already have on the computer. That’s where I’m hitting issues. If I had the VHS tape itself the. I’d use my upscaler that I physically have. But all I have is the digital media in 480p resolution.

I tried DVDfab Enlarger AI as it claims to be a good upscaler with AI being able to recreate the image. It used all 11GB of my GTX 2080TI, half of my CPU (which is 8 cores and 16 threads @ 5.00Ghz) and took 30 mins to render out 480p to 1080p. Video results where worst then just resizing in PD16. Gonna try another program I found that claims to do AI upscaling, but I got a feeling it may be the same software but rebranded.

Quote Your TV is the best upscaler in your house.
If you have captured VHS compressed, then upscale compress, then edit in PD and perhaps compress.
That is Satan's workflow.

these video were captured by someone else and sent to me. Now if I was doing them off as DVD, I probably do them off at the resolution they are in. But I’m doing these digitally. Unfortunately a computer monitor does not upscale, if it does then it not to good at it lol. The original on a 4K monitor is good at its native size, which is about 2 inches by 3 inches on the screen lol.

I really don’t want choppy looking video and I want to at least get 1080p quality if I can.

Now I know when I do old game systems like the N64, I can upscale the video with a physical $30 upscaler that takes the composite input, upscale to 720p and 1080p and spits the video out by HDMI. I used this before with a HDMI recorder with great results, it works so well that I been thinking of using my regular VCR instead of the VCR to PC by ION that I own. It just to bad that PD doesn’t recognize my HDMI recorder like OBS Studio does so I can’t record directly within PD, in fact it’s the only video recording software I have used that doesn’t pickup my HDMI recorder which is just a simple EzCap285.

But all in all, the issue is still the same, I got video that on the computer which needs to be upscaled for computer use. So being able to see the video in a higher resolution would be a great help. I was willing to pay for the upscale software till I noticed the end results were no better then just resizing in PD.

I been doing some old VHS videos and I'd like to upscale them to 1080p. I tried a trial product by a company as there AI based program came up as 1st on google; but it software made it seem like it was doing a lot of work to "upscale" the video. Took a half hour to do it, and used up 90% of my GTX 2080ti in overclocked mode to give me a final product that didn't look no better then if I just resize the video in my PD16 (Which take less time to render using less hardware).

I'd like to upscale the video and then work on it within PD16. One of these days a upscaler may get added to PD, and I'll upgrade on that day as most of my video content is old school footage.

Hey, I got a EzCap284 HD, I want to do so live recording directly into a project.

So I connect it by USB and installed the PC drivers from ezcap website for it.

When I go to capture in PD16 it doesn't show up. Anyway to fix this.

I used OBS Studio and it picks it up right away as IT9910 Capture Filter (HD) and works flawlessly. So I am guessing this is an interal issue with in PowerDirector

EDIT: Thanks Jeff, worked like a charm. Gonna use my key now to make sure it works.

EDIT: Everything works and I now have YouCam 7 registered to my account.

Thank you again Jeff for your help and support.

I did a survey today for cyberlink and they gave me a product key for YouCam 7, however the download link from cyberlink failed.

I figured I'd download the trial and activate it but there only YouCam 8 on the site, but the the product key was for YouCam 7 only for doing cyberlink survey.

This is the link provided to me that is suppose to be active untill May 31, 2019

Moderation note:
<Link removed>

^ --- But it doesn't work.

It been 8 days, A support ticket was made (CS001846490). Not a single respone, not one bit of help.

Does CyberLink normally have such bad support ???

Quote

Quote Found the problem. Patch would not complete because I uninstalled WaveEditor at an earlier date. Reinstalled everthing with WaveEditor included and patch worked fine.

Quote This is not complicated. The file downloads automatically from the update menu. 90% throught the installation of the patch it gives a message that it cannot extract file Data1.7z. Windows defender is off. No other antivirus installed. That's it.

Quote

Quote will this issue gone after you restart PC? it seems fine on my laptop.

no. and i re downloaded it again. This is not a problem with my computer.

There is no Dec 20 update to Power2Go 11. The last one is from Dec 13.
That is a pretty definitive statement.
I'm not having any problems with my desktop either so... .Where did you get the file ID as "Data1.7z" .
Did you look at the notes for this update and do you fit into one of these categories?
" Note: This patch is not for Power2Go OEM versions, or versions purchased via Steam.
Adds support on Windows 10 Fall Creators Update platforms.
Resolves the issue that certain YouTube videos cannot back up in 2160P (4K) quality.
Minor bug fixes."
Did you have your antivirus and/or registry cleaner active? It sounds like you have a zipped file - do you? Where did you get it?

I got the same issue, the lastest update stops at 90% with an error saying it can't extract the data from Data1.7z.

I have WaveEditor installed and it works. I have turned of my virus scanner. I own the full suite from CyberLink, bought it from the website here. So I think it's a corrupt file.

Quote

Quote Hi Elochai,

which windows 10 ADK did you install? win10 1607 or win10 1703?

1703

1703 means the ADK was released in 2017/Mar. which is quite new for now.

Maybe 1607 is safer, it was released on 2016/Jul.

I can try the 1607 but is there an uninstall program for the 1703 version or will it be in add/remove programs ?

Quote Hi Elochai,

which windows 10 ADK did you install? win10 1607 or win10 1703?

1703

I'm trying to use the feature to burn a system recovery disk for windows 10. I did the installion of the windows program that was requested by power2go using the link provided by power2go manual page for creating a system recovery disk. After I start the power2go utility for creating the restore disk everything seems to be fine and the DVD starts burning. Then it will quit in the middle of burning the 1st disk and says it encounter a read image error.

any ideas why I keep getting this error and how I can fix it. I don't really want to lose more DVDs to failed burns.

thanks,