GuildedTest part2: re-creating a similar gradient pdhowler 2026-09-22 | GuildedTest part2: re-creating a similar gradient
Local AI Translator - Batch translation progress late November pdhowler 2024-11-26 | this is a brief status update, work in progress, not released yet as of Nov 26. Just to show some of the great new features being added. WHen you have many text files you want translated, point the folder at it and go. The quality and speed of translations will depend on the model you choose to use, the speed of the system (RAM, CPU, GPU...) . The translation is local, doesn't need access to the internet. This is meant for translations that need high level of privacy.The software comes from MyDutchPal.com and uses the free Open-Source Ollama framework (ollama.com). You get to choose for free amongst many models. Some models are better, bigger, slower, others are faster, leaner. You'll find some models that are well suited for your needs.Stay tuned for more videos that will show comparisons of speeds observed with various models (llama 3.1, 3.2, qwen 2.5,...) of various sizes, from 0.5b to 20-30b - when you need it very very fast, you can pick a smaller model. When you prefer best quality of translation, go with the larger models, even if they are a little slower.
Status Update - mid November 2024 pdhowler 2024-11-18 | a few words about recent developments, new features etc...
Local AI Translator now supports Batch Translation of text files pdhowler 2024-11-07 | The Local translator from MyDutchPal.com can translate locally, using models found at ollama.com such as llama 3.1 or 3.2 models. There are options for translating text and content from several file formats.In this video we show some of the previous available methods for translating text, and introduce the new Batch translation capability.In this firstr release of Batch translation, it supports Text files. Imagine having a folder with hundreds, even thousands of text files. One by one they will get translated. This is intended for use in eDiscovery and forensics translations. The translations are done locally, no internet needed. You can disconnect the system from the Wifi or other network connections.In this example we run it on a laptop. 13th gen i9 with 20 CPU cores7-8 cores in use during translationGPU is NVIDIA RTX 4050RAM is DDR5 at 6400 MT/s
wipeout01s pdhowler 2024-10-27 | just a screen capture of a few brush strokes in watercolor FX standalone, with a starting image or "Wipe Out"
introducing watercolor FX update and standalone editions pdhowler 2024-10-27 | Only a few days ago we released the first as a plug-in. If you have PD Howler 2025, then this will be in clouded. If you don't use PD Howler but are interested in this, then you may be interested in the new standalone edition.It is now available from the author pdhowler.comIt is also available through thebest3d.com/salesYou can see more details about it at thebest3d.com/newsletters/207Happy Halloween. We hope this paint effect is not too spooky !
Circles 103 - how to use emf files from a vector program for circles in Howler pdhowler 2024-10-07 | There's one tool that in PD Howler can load Enhanced Metafile files (.emf).You can find circles and other shapes incluyded with Howler but you might also want to make your own, perhaps for other types of circles, thinner, thicker, filled, with dashed/dotted lines etc...This will show one way to do it with Inscape
circles102 - more ways to do something with circles pdhowler 2024-10-06 | there are more ways to create circular shapes with PD Howler.... These might trigger some ideas. I hope it inspires.
circles101 - how to draw a circle in a paint program pdhowler 2024-10-05 | Project Dogwaffle's PD Howler is a paint program. It does have some tools for more than brush strokes and painting. It does have rectangular tools and ovals. They either fill the designated area with the current fill settings, or they draw a hollow oval, with the current brush and its dozens of parameters, which could include dryout, random position, pressure based changes in size, even music/sound input to affect its appearance and modulation by another 'background' image. You can draw more than a circle with the oval tool.But can you draw a circle?This is not a simple "Use the Shift key" trick. But there are many other ways to create a seemingly perfect circle. Within reason.Let's explore a few. There will be more.
Watercolor_FX - quickstart installation short version pdhowler 2024-10-03 | a shorter version of the installation and first dibble dabbles
watercolor FX plugin for PD Howler 2025 - installation pdhowler 2024-10-02 | Here's a few tips and quick walk through, on how to install this new plugin for Howler: Watercolor FXAssuming you already have Howler installed, you can add this plugin:1. Turn off Howler session if currently running2. Copy the 2 files of the new plugin into the Howler installation folder3. Restart HowlerTo find and run the plugin, go to: Filter - All plugins - Filters - WatercolorHappy splish-splashing.Typo in credits for music:Jose (not Juse) Luis Suazo, aka eMaya - Mayan music for the Electronic age.
fast neural machine translation on premise with Systran pdhowler 2024-10-02 | This shows NMT powered by Systran Translation Server installed in a Hyper-V VM on WIndows 11 Laptop. Even with just 8 CPU cores you can get very respectable performance. Ideally 16 cores or more,. 12 GB RAM as a minimum. We've seen running it even on older PCs with DDR3 or DDR4 RAM at 2666 - 4000 MHz. The faster the RAM the better. If you have more RAM that also helps but it's not just the quantity of it, but also the speed of it. In this example we're on a Lenovo laptop with 32 GB of RAM but only about 16 GB dedicated to the Virtual Machine. The memory is very fast at 6400 MHz DDR5.This example had just 1 instance of each engine, one for English to French, another for English to Spanish. When both wre translating simultuneously, 16 cores get busy (8 per instance). In this simple demo however we only did the translations in a serial way, so only 8 cores are used, one engine at a time looping through 1000+ text files.The text is from FBI interrogations during the Chris Watts murder investigation. It was produced via ASR from the available video recordings.
Blender to Howler - running girl image sequemce2 pdhowler 2024-10-02 | I've had some trouble figuring out how to save a rendered animation from Blender, to then use it in PD Howler. I had assumed there would be a Save or export option. Turns out it's not the case. The output config parameters is where you define the desired format that will be used. Render options is one thing, but Output options is what creates the necessary files, in the desired format. Yes you can select PNG and it will save to that format for single image but also for Animation, rendering to PNG image sequence. Those you can then load into the Howler animation or animated brush.. Here I was using Adobe MIXAMO with a character running in place, rendered over a 22-frame loop. I rendered it out. The folder I designated in Output was C:\tmpThe files automagically get rendered there so all I need with Howler is to look for them. No need to 'Export' or Save them from a File or other menu. Now I do like it.However, I have yet to find the AVI option that will support Lagarith codec which is lossless and supports RGBA. I see an option that's powered with ffmpeg and it has a bunch of formats and for AVI shows a bunch of codecs, but I don't see Lagarith. Bummer. Oh well, it'll work at least with the Image sequences.I forgot to mention: music by John Somnitz, aka DJ Dartagnan. Please support the regional MS (Multiple Sclerosis) Society, perhaps for upstate New York. Thank you!
Dans Watercolor Simulation POC pdhowler 2024-09-20 | proof of concept, work in progress. Something coming to Howler 2025. Watercolor simulation
Howler 2025 - Whats New? pdhowler 2024-08-25 | here's a walk through into the mostr important new features for the initial release of Howler 2025. Late August 2024.A labor of love and passion, in time for Labor day.
TL03 Speech to Text - offline - with Llama3.1 and other models from ollama.com pdhowler 2024-08-23 | powered by llama3.1 through ollama.com and integrated with a Python program created by Terence Wilson of MyDutchPal.com - this does it off the internetYou may already have several models from ollama.com, and some are better than others. It keeps getting better. The tool developed by Terence runs locally, you can have the desktop or laptop offline. Great also for transcribing and translating sensitive material. (legal, wiretap, surveillance, interrogations, OSINT,......) or just about anything with audio, even audio tracks of video.More on this at thebest3d.com ... The 2nd "D" is for Digital Dialogs
TL02 Backtranslation with Llama3.1 and qwen2 pdhowler 2024-08-23 | a feature that will be useful when you don't know the language and you want to compare the original to a back translation to see if there's a good chance it'll be good enough. Not a guarantee, but usable in many cases when you don't have access to a human certified translator that knows the two languages and the subject matter too.
TL01-Rag Intro 240808Rev01 FHD24 w/Music pdhowler 2024-08-23 | a very short version of RAG explained from a user perspective with tools by Terence Wilson
August16 Transporter11 - one shy of Master Chief pdhowler 2024-08-17 | If you know your Halo and Master Chief, the number 117 has a special meaning. This launch had 116 payloads. Close but no dice. Still awesome.
Newstuff - mid July 2024 updates, filters, SFX coolness pdhowler 2024-07-12 | A walk through a few recently added features in PD Howler - get it at pdhowler.com or thebest3d.comSome recent filters are created with help of AI such as Copilot
LLM Ollama - Free Translations on premises, off the cloud pdhowler 2024-07-01 | when dealing with sensitive material, on-premises or even better, on device, can be a good cheap way to get it done. Many LLM can be deployed locally and used without internet - for free.You can also add interfaces to it or write code to integrate. In fact the LLM can generate the code to your specifications too. (Dan Ritchie at pdhowler.com has been creating new VB scripts for Howler with Copilot, for example).Just a few steps and clicks and you'll be learning and working with Kazakh.... what's not to like, Borat?
How I adjust audio after recording pdhowler 2024-07-01 | typically using Power Director, sometimes other tools, I usually need to adjust the volume and sometimes do some noise reduction
VB script: galactic2 pdhowler 2024-06-30 | script was created with GUI by Bing Copilot, then several variants of the rendering combined.Look for more samples and details at thebest3d.com/dogwaffle/newsletter for example:thebest3d.com/dogwaffle/newsletter/205-SummerScriptingDan's Patreon updated API doc related to these new scripts is here:pdhowler.com/API_And_Scripting.htmWe also recommend OpenCV - open source Computer Vision - if you do any integration of OpenCV with Dogwaffle please share
Lightning Bolts in digital painting pdhowler 2024-06-19 | here's another VB script coming soon to Dogwaffle, created with the help of AI: Bing Copilot. This one lets you create various lightning bolts. You can then take them in for extra processing, light diffusion, globs, other details that make them realistic.
how to waste time NOT playing video games pdhowler 2024-06-15 | if you start this game you have 4 minutes to do your homework, tidy up your room or take out the trash, water the lawn, feed the cats, walk the dogs, fetch the mail.... order pizza ...etc etc...before it actually starts
Galaxy script pdhowler 2024-06-11 | getting a galaxy script (vb) put together by ai - this one Bing Copilot, using it in PD Howler
H24 Videotools08 part6- blue screen : compositing 2 videos pdhowler 2024-05-26 | one of the videos from the newest tutorial collectionthe complete set is available at thebest3d.com/sales
H24 VideoTools02- Preamble: (ab)using Anime Studio for easy video conversion pdhowler 2024-05-11 | Here's a preamble, with lots of admiration for an intuitive tool from MoHo. I was known as Moho,, then AnimeStudio, then back to Moho. It only got better. And one of the ways I love to use it is to convert my mp4 videos such as from my smartphone to an image sequence or better yet to make them into an AVI with the lossless Lagarith codec. Works great for PD Howler's import of AVI files when you can't import the MP4.
Howler 2024 VideoTools - the Teaser pdhowler 2024-05-10 | Here's a sneak preview of some of the tricks and tips we'll explore in an upcoming new tutorial collection focused on working with video.- we'll create an animated sky for looping backgrounds- we'll use some drone footage (i.e. footage of a drone)- we'll try to extract its mask for use in separating it from the background- we'll combine the RGB stored animation with the stored selection mask- we'll use it as an animated brush in the brush keyframer- we'll see a few other related tricks, with the brush timeline editorHope you can join!
wooden picnic table - kudos to poly haven pdhowler 2024-04-27 | wooden picnic table - kudos to poly haven for making these model possible and available for personal and commercial use.You can load it in Blender and re-export to Wavefront OBJ, then load that into PD Howler. I played with some parameters, both in the interface of Howler and also in the .mtl file that came out of the OBJ export in Blender. Rendered a bunch of adjanced rotated views, collected into an animbrush. Then rendered that with the brush keyframer. Then used the motion prediction module for motion-estimated smooth animation. A little bit of post, all in Howler.The PBR renderer is cool but I wouldn't say it's complete yet, we don't do raytracing yet so no shadows. If you have baked ambient occlusion that would work though for extra shadingKudos and big shoutout to polyhaven.com !!!
Gradients With Transparency pdhowler 2024-04-21 | the gradient tool carries RGB channels, and it can also use alpha. Use it to open sharp or soft holes in the gradient. Alpha is used for opacity control, the opposite of transparency.
Tablet recording - using a new low cost pen tablet with Project Dogwaffle pdhowler 2024-04-15 | Did you hear under $20 ? or was that $120 ? no, under $20
teaser - from chocolate coffee to Mars and beyond pdhowler 2024-04-13 | I'm working on a few tutorials for this series, starting with ordinary frothed coffee (and other pictures) to use as elecvation map for fantasy landscapes
Working with Stabilized Video - teaser intro pdhowler 2024-03-24 | In this tutorial series (in the making as of this upload) we'll take a look at a number of tools found in PD Howler for working with snippets of video. We start with a short and very shaky video of 55 frames, that's around 2 seconds. Then we stability it, fix some bad frames, render something else into it, and restore the original motion. Finally we use the motion estimated interpolation to make it last longer, 5-6 seconds is what you needed it to be? No problem.
MyMedia13b - sample animated Sky filter, with slight post pdhowler 2024-03-12 | all done in minutes (or seconds depending on the speed of your system and dimensions of the video) on PD Howler 2024
intro to new tutorial: meteor shower pdhowler 2024-02-18 | a multi-part tutorial is coming soon, this is an intro. It will be available at thebest3d.com/sales
