Halide Developer Takes a Deep Dive Into iPhone XS Camera and Skin Smoothing Rumors

[buststyles]

That’s quite true isn’t it... plus the fact it happens with faces only. This article needs to be scrapped.

Question is this smoothing only happening with the front facing camera. From the posts I've read this appears to be the case. What else would you take a picture of with the front facing camera. The article also posts a side by side of shots from apple in different lighting conditions that shows less softening in better light.

 

[Diatribe]

Question is this smoothing only happening with the front facing camera. From the posts I've read this appears to be the case. What else would you take a picture of with the front facing camera. The article also posts a side by side of shots from apple in different lighting conditions that shows less softening in better light.

"Smoothing" is actually noise reduction and it is happening on both cameras, just more prominently on the front camera due to its worse light sensitivity. It is purely software as you can see in the concert shot on the Halide blog post. The Halide RAW is pretty darn good while the iPhone shot is unusable.

 

[H3LL5P4WN]

Cool. I'll give it a go.

The. Most intriguing thing about the app is some of the features it seems to have that are exclusive (natively) on Xs models. Have you seen or done any comparisons between halide on an X and stock xs shots for photo correction?

No, but actually that's a hell of an idea since my GF has my X. What settings would you like me to test?

Also FYI, at the time of this reply, Halide has not updated with the new Xs features.

 

[Unimpossible]

You’re assuming Apple doesn’t know about the issue or somehow needs to be pressured into improving the pictures. Rather, I think Apple is well aware of the issues and limitations of the XS’s v1.0 computational photography algorithms. No doubt they’ve been working on improvements since 12.0 was frozen.

And I’m sure potential customers do care, and they’re not who I’m referring to when I talk about people who really don’t care whether or not the problem gets fixed.

Right. So you’re saying your feedback-link is just as futile as media awareness, because Apple is already know about this issue?
Apple are aware of a lot of issues, from feedback-links and media, that they will never assess and/or improve.
But media «pressure» is quite effective, and more so than that feedback-link.

 

[PickUrPoison]

Right. So you’re saying your feedback-link is just as futile as media awareness, because Apple is already know about this issue?
Apple are aware of a lot of issues, from feedback-links and media, that they will never assess and/or improve.
But media «pressure» is quite effective, and more so than that feedback-link.

Media pressure can surely be effective, but I don’t think a bunch of hysterical ranting about “beauty mode” in a YouTube vid comment section or a forum thread is as effective as you might think. At least using the feedback link allows Apple to quantify the magnitude of the suggestions/ complaints.

 

[Unimpossible]

Media pressure can surely be effective, but I don’t think a bunch of hysterical ranting about “beauty mode” in a YouTube vid comment section or a forum thread is as effective as you might think. At least using the feedback link allows Apple to quantify the magnitude of the suggestions/ complaints.

Hysterical ranting? Where?
And that feedback link doesn’t quantify anything.
The huge cases Apple has had issues with are all by media attention, not by that feedback link.

 

[PickUrPoison]

Hysterical ranting? Where?
And that feedback link doesn’t quantify anything.
The huge cases Apple has had issues with are all by media attention, not by that feedback link.

Well Apple knows exactly how many people are reporting via the feedback link, it’s trivial for them to know when there’s a real issue. I don’t know what your last sentence means.

 

[Feenician]

Seems youtuber Jonathan Morris caught rather a large number of Android fans out. He posted a couple of selfies on Instagram (https://www.instagram.com/p/Bokj_AwhKR6/?utm_source=ig_share_sheet&igshid=1nhpjwfmooq6l and https://www.instagram.com/p/BokNrIrh-yY/?utm_source=ig_share_sheet&igshid=1w89u49sh9sge), saying they were from a Pixel 2. Got many responses saying they were great, way better than XS, didn’t suffer from “Beautygate” etc.

Problem is he later made a video revealing they were actually shot on XS

An interesting demonstration of how people’s personal biases affect their judgement

 

[lordofthereef]

No, but actually that's a hell of an idea since my GF has my X. What settings would you like me to test?

Also FYI, at the time of this reply, Halide has not updated with the new Xs features.

Truthfully don't even know lol. I just know the xs supposedly as a srockfragure will take multiple shots in one photo and then use the best parts of each to intelligently create the best image.

The largest side by side comparison difference dive seen are when the light source is behind the subject. The x (stock) seems overblown. The xs corrects this greatly.

 

[Unimpossible]

Well Apple knows exactly how many people are reporting via the feedback link, it’s trivial for them to know when there’s a real issue. I don’t know what your last sentence means.

Of course Apple knows how many uses the feedback link. That still doesn’t say anything about the magnitude of the issue, as there are much more people addressing the issue in social media and on forums.
As to you not understanding a sentence, I’m not surprised.

 

[PickUrPoison]

Of course Apple knows how many uses the feedback link. That still doesn’t say anything about the magnitude of the issue, as there are much more people addressing the issue in social media and on forums.
As to you not understanding a sentence, I’m not surprised.

Apple can determine the magnitude of various problems based on the the reporting received through their feedback system (and of course through tech support) irrespective of social media/forums.

A relatively small issue affecting a small number of users can be blown out of all proportion on social media, so it really doesn’t represent the magnitude of the actual problem very well.