On Foveon X3 Quattro...

Started Feb 10, 2014 | Discussions thread
RussellInCincinnati Veteran Member • Posts: 3,201
you're contentiously searching for negatives

Erik Magnuson wrote:

RussellInCincinnati wrote:

Just because you don't get "any more resolution" out of upsizing doesn't mean you don't get anything from upsizing. For example, sometimes you get a little more bang-for-the-byte with more-lossy JPEG compression of higher-resolution images with a lower quality JPEG output setting,

This depends on very specific images and compression ratios. As a general case, it's not worth it. Remember you equated the value of upsizing with the value of offering downsizing.

I didn't say and you are contentious to imply, that I am foolish enough to have said that upsizing has the identical value to just as many people as downsizing. You even reproduced above my main point that there is some value to upsizing for some people...I can't imagine anyone finding that simple truth so repellent that they think it's in poor taste or somehow negative for Sigma to offer an upsizing option.

E.g. JPEG output at 90% quality and 40 megapixels could easily be smaller than JPEG output at 100% quality and 20 megapixels.

Yes, you can contrive such cases.

Somewhat like people contriving weird cases for downsizing in-camera JPEGs when memory cards are so huge, buying expensive equipment but never learning to post-process, trying to find something negative about a certain JPEG output scaling option, etc.

Are the algorithms Sigma is using favorable for this? Do they occur often enough to justify the mode? (Answer with previous Sigma implementations was no.)

Oh well if Sigma did some kind of loss-of-detail downsizing poorly (that nobody who's interested in the unique fine detail quality so much as to buy a Sigma would ever use anyway) in the past, then it's surely time before the cam is out, to wring our hands that some new probably-no-loss-of-resolution option might not be what someone wants.

Also if someone had a workflow that wanted real high res images as input, it is an optional convenience for a Sigma Quattro to offer that high res straight out of camera.

They are not "real high res" just upsized.

Irrelevant to the fact that some consumers of images want a certain minimum pixel dimension in submitted images. And thus some image creators might want just such huge JPEGs as a Sigma output option. Who the heck knows what the real "native resolution" was, of any particular Bayer sensor camera anyway? My Bayer camera only has 4 megapixel resolution in real red and real blue areas for example.

You will struggle to find a workflow that wants the increased dimensions and yet doesn't want control over the algorithm used to produce them.

By this logic nobody using expensive cameras ever wants JPEG output (certainly it's not my interest). But it is undeniable that many a user of real expensive cameras wants JPEGs. And if those people incontrovertibly exist, it's not a shock to one's senses to think of firmware offering in-camera upscaling for those JPEGs. This option is so trivial and unobtrusive that it's feeling a waste of time to discuss one's level of interest in it.

Also it is not 100% clear to me yet that the lower pixels are perfectly under the topmost layer pixel areas. Some games with guessing higher-than-20-megapixel resolution details are conceivable (not saying likely) if there are offsets between Quattro layers.

Even if there are such possibilities, they are extremely unlikely to be supported for in-camera JPEGs..

Interesting. Though we do see a new CPU and new battery power available for the Quattro cameras. Thus a little early to say how industrious the new firmware is "likely" to be. Now you could be dead on right and all the new cams have is larger batteries, smaller raw files, and no "better" image-quality-developing firmware.

Not sure why you mention this, since we can probably agree that across all cameras, people do JPEG's usually for convenience,

Extra large JPEGs are not convenient.

You appear to be purposely inflammatory. (a) have already described how larger-pixel JPEGs might be conveniently smaller files depending on quality options. (b) Obviously you have enough wit to know that it's more convenient to have the camera produce larger JPEGS if you want larger JPEGs at the end of the day, than it is to have to do your own up-scaling in post-processing. Since the whole point of JPEGs is what the cam do without post-processing. Not sure why you are pretending that "extra" large JPEGS "are not convenient" when it's not rocket science to acknowledge that someone might want such (possibly smaller than 20 meg even) dimensioned images to pop out of the cam with no work, as an option.

Yes, really. Some users/print shops want 600 dots per inch.

So this will help all of those users/print shops who want to print 12.8" x 5.3" (or exact multiples). You only get one large size option in a couple of different aspect rations and it doesn't match anything particularly convenient.

Maybe just a bit like how great and fine people feel--who hand-hold their 36 megapixel Bayer cameras, with questionable technique, focus and lens quality, looking at fluid scenes, where nobody's ever going to see anywhere near a sharply rendered 36 megapixels.

Um, Quattro is not immune to this. If it's really equivalent to a 40MP mosiac sensor, then it is just as subject to all of those same technique issues. So now you have the even more contrived scenario of someone who cares enough to shoot such a camera with careful technique yet is happy to print/use in-camera upsized JPEGs. My point is that this subset of users is extremely small.

And the importance of that point is? It feels like a point of someone searching for negatives or contention or criticism of some camera maker, or of other posters' intelligence...related to a camera that we don't even know what anyone can do with it yet.

Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow