Hayley Williams For M.a.c Untoccodiglamour



Here are the hot products making up the Hayley Williams M∙A∙C collection, available exclusively online at maccosmetics.co.uk from 9th April 2013. Hayley Williams of Paramore teams up with M.A.C. For a four-piece cosmetics collection.

This review is long overdue but it’s better late than never, right? 🙂 Alright, let’s go!!! Last Spring, M.A.C. Launched an exclusive online colour collection inspired by rad front woman, Hayley Williams, from platinum-selling band, Paramore. Having fallin’ in love with Williams’ incredible voice, her vibrant flame orange hair and wicked makeup, I felt I’d been waiting for this partnership since I watched their “Crushcrushcrush” MV back in 2007.

I was beyond stoked. The 4-piece colour collection ranged from orange pop to shimmery coral–variations of Williams’ “signature” orange shade.

Hayley williams for m.a.c glamour

With names like “Sounds Like Noise” and “Daydreaming”, Williams also incorporated Paramore by naming the products after a song lyric or song title which, as a fan, I thought was a nice detail. The collection consisted of frosted mid-tone coral eyeshadow “Daydreaming”, matte bright true orange lipstick “Sounds Like Noise”, bright true orange nail lacquer “Riot Gear”, and soft candlelight beige with multi-dimensional shimmer mineralize skinfinish “Lightscapade”. I wasn’t as interested in the mineralize skinfinish but am kicking myself now for not picking it up.

For your viewing pleasure, here are some shots I took: The eyeshadow has a wonderful pigment and texture true to M.A.C. It has an interesting reddish orange tone that reminds me of a robin’s breast feathers. This shadow goes well alone or with rich browns. The matte lipstick has a creamy opaque finish. I prefer to sheer it out with Smith’s Rosebud Salve but it’s an absolutely beautiful shade straight from the bullet.

It reminded me of the Manic Panic hair color shade “Tiger Lily” which, I believe, is one Williams once wore. The nail lacquer is, by far, my favorite of the three items I chose. Unlike most neons, the first coat was streak-free. I was very happy with it’s high-gloss finish which hardly chipped (below are photos of my fresh mani and the same mani 5 days later). Since I work with my hands and am constantly removing makeup from them, this was a major plus.

M.A.C.’s nail lacquer formula also contains conditioners and UV protection which I absolutely love!!! Although I was satisfied with the quality of the products, I wanted more!!! I hope M.A.C. And Hayley Williams will partner up again in the future and create a larger collection with maybe an eyeshadow quad palette, a blusher or multiple lip colors. I’m holding my breath and I got my fingers crossed!!!! Hope you found this review informative.

I guess graphics card prices aren't going down anytime soon;) copy-pasted from OP's reply: Here's the settings I used. About 10-11k dataB images from various video interviews and google image downloads that have been sifted through. Trained against the pre-trained model from the other vid I posted, which had already trained for about a week on a 1080 ti, getting that loss value down to 0.013. For this clip, I let it sit for a couple days and the loss value was down to 0.09 when I stopped it and ran the conversion. Using Fakeapp 1.1. Batch size: 256, save period: 100, layers: 8, nodes: 1024, mem ratio: default, GPU growth: false. For combine, seamless: false, blur: 10, kernel: 8.

Other tricks: The converted video was 1080p. Downscaling it to 720p helps blend the original vid with the overlaid face. At original resolution, the original vid looks too sharp. A decent amount of compression on the vid helps too as it blends the overlaid face into the original video.

Also, with face extraction, don't downscale the images with LIR first - the extractor can handle 1080p source PNG files, at least on my card. Upconverting to 60fps seemed to help with details in the face as it was easier to track while she er, bounces around. As others have mentioned, finding a decently matched body helps. Kasey Chase isn't perfect, but similarly skinny to Emma I guess. Some of the vids someone else has done using Tori Black look great too - the face shape on Tori works well with Emma. Another thing that seems to help the illusion - hair.

Hayley Williams For M.a.c Glamour

Emma Watson always has full, wavy/bushy hair. I chose this clip for the experiment because Kasey has a haircut with thickness/fullness similar to Emma's. In a lot of Kasey Chase's clips, she has flat hair, which I don't think would have worked very well. That's another reason I think Tori Black seems to work well as an Emma surrogate.

Another trick I'm playing with now that I'm not sure works or not is spot training. I notice the two vids I've done have spots that suck, where the model didn't train as fast as the rest of the faces. I'm going back and only putting the frames that got a weak conversion in the dataA aligned folder, then training the model more against it. Another conversion and those frames should be better. If they are just copied over the rest of the converted frames and overwrite the crappy ones, then stitch the video together, that may solve the problem where some areas of the vid get really blurry/weird.

For example, the full length clip that this test vid came from, there are some blowjob parts. I'm currently training the model against just those frames. It's going really slowly! I believe because it just hasn't been trained on that type of image yet. I have faith it will learn with enough time. I supplemented my dataB aligned set with a good amount of side views of Emma to try and feed better data to the training. We'll see what happens.

/ / / / / / / / / / / / / / / / / / / / / - + + + - Copyright © 2010-2014 Tinyboard Development Group Copyright © 2012-2014 vichan-devel Copyright © 2013-2018 N.T. Technology, Inc. Based on sources from Fredrick Brennan's 'Infinity Development Group' by All posts on 8chan are the responsibility of the individual poster and not the administration of 8chan, pursuant to 47 U.S.C.

We have not been served any secret court orders and are not under any gag orders. To file a DMCA takedown notice, please email. In the interest of transparency, notices of claimed infringement will be posted on.

This entry was posted on 06.10.2019.