@sakuramochi55@sakurajima.moe avatar

sakuramochi55

@sakuramochi55@sakurajima.moe

Academic in training. I like a lot of media, whether it's music, drama series or anime, or even printed book.

This profile is from a federated server and may be incomplete. Browse more on the original instance.

anianimalsmoe, to random
@anianimalsmoe@sakurajima.moe avatar

My workflow for Anime Animal processing is:

  • Anime watching: VLC (supports all damn formats)
  • Format standardization: Handbrake (supports all damn formats)
  • Video trimming: Adobe Premiere Pro (for precision)
  • Video/Image Watermark + Thumbnail: Adobe After Effects (for batch processing automation)
  • Animal Thumbnail: Adobe Photoshop (for blur and AI Remove tools)

I wonder if there's any good FOSS stuff to replace the Adobe Suite in there. I was trained in this stuff as a student, but paying for it without the Education discount's been quite a pain. :leafeonmoney:

Any good suggestions?

sakuramochi55,
@sakuramochi55@sakurajima.moe avatar

@anianimalsmoe You could use Avidemux (AppImage-based) to extract frames from video files, and use Kdenlive (either AppImage or distro) to process the watermarking thing. But you could also use GIMP to add watermarking if it's a lesser volume.
But these may not be available for Silicon Macs, but rather on Linux distros.

sakuramochi55,
@sakuramochi55@sakurajima.moe avatar

@anianimalsmoe Kdenlive also has this speech-to-text function as well but that one relies on Python and either VOSK or the vanilla version of openai-whisper (which doesn't have very accurate recognition), and if you're looking to create usable subs for niche series you have better options such as whisperx that can use better recognition models and phoneme alignment models. (Still might be limited to x86 Linux / CUDA)

sakuramochi55,
@sakuramochi55@sakurajima.moe avatar

@anianimalsmoe according to the manual you can generate scripts to use with batch rendering, but you could also use ImageMagick to add watermarks, provided you're backed up the files beforehand. There's this tutorial for it

sakuramochi55,
@sakuramochi55@sakurajima.moe avatar

@anianimalsmoe also, you can use VLC to save frames as well but I guess it's not as granular as with avidemux, which loads in every frame which then can be extracted individually.

chikorita157, to LycorisRecoil
@chikorita157@sakurajima.moe avatar

I guess I should watch the first stage page recording of Lycoris Recoil sometime in the future considering they are doing a stage play sequel. :bunhdthink:

https://www.animenewsnetwork.com/news/2024-01-15/lycoris-recoil-anime-gets-2nd-stage-play-in-june/.206483

sakuramochi55,
@sakuramochi55@sakurajima.moe avatar

@chikorita157 @yon have you tried using WhisperX for creating the subs?
I think it can create a good starting point which you can improve.

anianimalsmoe, to random
@anianimalsmoe@sakurajima.moe avatar

Someone was using alts to downvote all my new Reddit posts for some reason, talk about being petty. :blobfoxcatterrified:

I literally make one post and it gets 4 downvotes. I delete it and rephrase it to something better, and it gets downvoted to oblivion again.

I'm just not gonna bother offering advice there anymore.

sakuramochi55,
@sakuramochi55@sakurajima.moe avatar

@anianimalsmoe I had around 47m karma there, but since I was always been using RIF as my go-to client, I figured I wouldn't install their bloated and notoriously unreliable mobile app, but I also didn't want to see my account reanimated from the dead, all I did was to save all my posts from the site, then silently delete my account. Since then I already moved to a Lemmy instance instead.

sakuramochi55, to Anime
@sakuramochi55@sakurajima.moe avatar

I've just finished watching the Tsukuyomi: Moon Phase anime series, after I've discovered its soundtrack album.

Since it's a vintage series (it's from 2004), most of the streaming sites that would have it in their catalog showed it as not available, I actually found it using a local source which was very helpful, and I just had to source subs for it.

As for the series, I really liked it, mostly because it had a quite good story and the gags have been very funny, especially the metal pan ones.

I also liked the fact that the intro sequence of the show changes at some episodes.
And I also loved its soundtrack, which was the main reason for seeking it out.

sakuramochi55,
@sakuramochi55@sakurajima.moe avatar

@anianimalsmoe well I originally got to know the series from the Neko Mimi Mode song then when I was searching for more about the series, I came across its OST collection album, and when checking it out, it further piqued my interest, but then I saw that it wasn't available from the source it should have been available based on surface level research.

But then I checked out deeper level sources that likely had it and I found a fansubbed version of it on a site that's local and is not sketchy. But because that site is optimized for quick load and mobile devices, it was only in 360p but it had a complete roll of it, and made adjustments to it to match the secondary sub I managed to find.
Because of this condition I didn't really check the animation quality but it seemed good. Its drawing style reminded me of the toaru series mixed with 90s anime style.

sakuramochi55, to fediverse
@sakuramochi55@sakurajima.moe avatar

So yeah, server blocking is a thing in as well, but they are called "Instance Mutes" and you "import" the blocking list via the textbox of the page, then applying the changes via the Save button.

anianimalsmoe, to random
@anianimalsmoe@sakurajima.moe avatar

deleted_by_author

    •
    sakuramochi55,
    @sakuramochi55@sakurajima.moe avatar

    @anianimalsmoe @chikorita157 Tachiyomi's J2K fork is likely a better one, because it also includes tracking support, like you can sync data with AniList and others.
    https://github.com/Jays2Kings/TachiyomiJ2K

    sakuramochi55, to fediverse
    @sakuramochi55@sakurajima.moe avatar

    Today's post:

    Nobody's born cool, except those blog owners and website owners who still offer RSS feeds on their sites.
    In these times, when birdsite is getting more and more degraded, RSS feeds are a godsend option for getting news articles into the .

    If you have a blog / site, you're likely to have an RSS feed ready to use, even if your browser doesn't display its recognizable logo in the address bar.
    The most simple test you can do to probe for a possible feed is to simply just append /feed to the end of a website / blog article.
    If the feed was found, you will be offered to download an .xml file, which is the RSS summary file, and it's likely that it's going to have a mix of letters and numbers as a file name, and might not even have an extension.

    A more advanced approach would be to right click on an empty space on the site and selecting View Page Source.
    While in this new window, you can search for rss, feed, rss+xml, atom, feed+json in hopes of finding one, but it is quite a rare sight to have one on modern sites.
    Another good measure would be to install an RSS discovery addon to your browser, which will return the feed indicator logo and likely will offer to copy the feed link to your clipboard.

    However, if you found a site that is entirely devoid of feeds, you might be able to use a generator to create a feed specific to the site, using RSSHub.

    sakuramochi55, to Wikipedia
    @sakuramochi55@sakurajima.moe avatar

    Today's :
    Did you know that there's a knowledge base operated by the Wikimedia Foundation, which acts as the central storage for all Wikipedia links?
    That site is Wikidata, and you can use it to retrieve information on certain subjects, which are mostly related to all existing articles present on the Wikipedias around the world.
    For more info, you can visit this site for a more detailed introduction:
    https://www.wikidata.org/wiki/Special:MyLanguage/Wikidata:Introduction
    You can also use the Query Builder if you're already familiar with the site and want to fetch info from the database:
    https://query.wikidata.org/querybuilder/?uselang=en

    sakuramochi55, to random
    @sakuramochi55@sakurajima.moe avatar

    Daily :
    If you have a video file but you don't have a subtitle for the video, and would like to create using local speech to text, you can use to generate subtitle for said video.
    You have two options for speech to text engines: VOSK and OpenAI Whisper.
    Both are using different types of models you can use, and CUDA computing or the good old CPU, and VOSK requires separate models for each language, and does not produce a formatted output, just raw text.
    Whisper is slightly more advanced, because it uses a multilingual model by default, which you can select to translate into English, from any language in the model and the output will be formatted normally, and acronyms like GPU and such are properly capitalized in the final text.
    But, as with anything, there's a catch: If you would want to utilize CUDA computing and the large model, you would need around 10 GB of VRAM, which isn't very common these days. However, you can always use the default option to use CPU compute, but that'll be around a couple hours, but in the case of a 24-min video, it'll be likely 40-50 mins to create the subtitle, which is a nice waiting game. However, once it's done, you will have a somewhat usable subtitle, exactly tied to the speech in the video. More details and info:
    https://docs.kdenlive.org/en/effects_and_compositions/speech_to_text.html
    https://kdenlive.org/en/download/

    sakuramochi55, to Anime
    @sakuramochi55@sakurajima.moe avatar


    Hello, I'm Jim, but some may call me Tim.
    I love and , and I'm an avid enjoyer of Japanese music, whether it's city pop, kawaii metal, or , or .
    I also like to watch drama series as well.
    I also like , and , , and .
    I also like to read actual printed books as well.
    I mostly used birdsite to get updates from the artists I listen to but I'm eager to change that to a much more open approach.
    (also, my profile pic is a gradient of the hanami dango's colors.)

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • JUstTest
  • kavyap
  • DreamBathrooms
  • thenastyranch
  • magazineikmin
  • tacticalgear
  • cubers
  • Youngstown
  • mdbf
  • slotface
  • rosin
  • osvaldo12
  • ngwrru68w68
  • GTA5RPClips
  • provamag3
  • InstantRegret
  • everett
  • Durango
  • cisconetworking
  • khanakhh
  • ethstaker
  • tester
  • anitta
  • Leos
  • normalnudes
  • modclub
  • megavids
  • lostlight
  • All magazines
    •