Ugly Hedgehog - Photography Forum
Home Active Topics Newest Pictures Search Login Register
Check out True Macro-Photography Forum section of our forum.
Posts for: Doc Barry
Page: <<prev 1 ... 5 6 7 8 9 10 11 12 13 ... 48 next>>
Nov 23, 2020 02:23:45   #
Ava'sPapa wrote:
I just ordered one from ebay (it was $95 with tax). Thanks for the info. Can't wait to get it.


I just ordered one from pmcsupplies.com with free shipping, no tax, plus a $4 discount Honey found. Total was $75.95. This will be a nice addition to my toys. Always a pain to set up lights, tripod and camera to photograph jewelry. And a soft box is also a bit of a pain.
Go to
Nov 21, 2020 00:14:09   #
Mark Sturtevant wrote:
I did not know that Apple purchased Lytro. And yes.. Hmmmm. They are maybe up to something.

The new Canon sensors are all very interesting. Even if their direct applications will be very specialized, I wonder if some of the technology will trickle down to us regular photographers.

Pardon my bad recall Mark. Google purchased Lytro for about $40 million and should show up in their Pixel cameras at some point.
Go to
Nov 19, 2020 16:21:31   #
CO wrote:
I know they've had the 250 megapixel APS-H sensor since 2015. There's a 120 megapixel sensor also. It won't be used in DSLR or mirrorless cameras. It will be used for surveillance, inspection, and scientific purposes.


Actually, as you mentioned, Canon announced they had a prototype 250 megapixel APS-H sensor back in 2015, but it was not available for purchase then. It was mid-2018 before they offered for sale (B2B only) the 125 megapixel sensor. In late October 2020, Canon offered the 250 megapixel version for sale, but still B2B only. Canon has made a prototype movie camera that had generated some nice promotional video. I expect they used one of their excellent UHDTV broadcast lenses.

Canon has indicated they expect this sensor will be used for specialty camera in the movie industry. Actually, this may make light-field photography practical for the movie industry. It will be, as they say, a game changer if it does the job. It has been tried before with smaller density sensors and lower processing power with marginal results. Many directors liked the promise of being able to take a scene and then in post adjust the area to be in focus and the depth of focus. The attempts at light-field cameras were not successful (Lytro). But note that Apple purchased Lytro and made it vanish. Hmmm ... the iPhone 12 Pro MAX is loaded with computational optics. Think about it.
Go to
Check out The Pampered Pets Corner section of our forum.
Nov 19, 2020 01:16:01   #
Canon has announced their new color and monochrome APS-H size CMOS sensors (29.4 mm by 18.9 mm) that offers 250 megapixels. The pixels are 1.5 um square which provides a spatial sampling frequency of 333 lp/mm. The readout rate is about 1.25 gigapixels per second or a frame rate of about 5 frames per second.

Canon cleverly included the ability to sample the image over a ROI (region of interest) which allows faster frame rates. For example, 4K video can be run at 30 fps and HD at 60 fps.

Lenses are bandlimited which means that even a perfect lens cannot propagate information above a spatial frequency given by 1 / (wavelength * F/#). Should one have a lens that is diffraction limited operating at say 500 nm and F/2.8, the cutoff spatial frequency is just over 700 lp/mm. The diffraction spot size for such a lens is 3.4 um (diameter). These sensors can make a beautiful imaging system that can mitigate spatial-phase vacillations and capture the extent of the information in the scene.

At its full data rate, it will fill a terabyte drive in about 6 minutes!

It will be interesting to see if and when Canon introduces this sensor in a professional photographer's camera. Model numbers are LI8020SAC and LI8020SAM.
Go to
Oct 28, 2020 15:36:21   #
ImageCreator wrote:
1. shows the original stitched pano image--183 mb size. The lake is lake Helen in Lassen Volcanic national park. Normal PP but no saturation. Lest one thinks its over saturated this is the way it came from the camera.
2. image 2 shows the addition of a sky--my sky--using Luminar 4 and then trying to bring the cloud reflections down into the lake. In hindsight I should have lightened this final image.
3. image 3 is a sky replacement using the new PS tool. Again, using the same sky as in the Luminar image--a sky I captured myself, not a preset sky. I then tried to copy the sky and move it into the lake for cloud reflections.
Is it perfect. Probably not as some will point out. But it does give a comparison of the two methods. One item of note. Due to the size of the original pano stitched image, processing in Luminar took a long time. Both methods took about the same amount of time to complete, except for the long processing time in Luminar. I thought it was easier in Luminar to bring the sky reflection into the lake.
If no reflections are needed in water I'd probably use the new PS tool.
1. shows the original stitched pano image--183 mb ... (show quote)



Nicely done. Luminar better. Need to revert the reflected sky.
Go to
Sep 26, 2020 11:52:14   #
E.L.. Shapiro wrote:
When photographers like or dislike any particular PRESET type portrait retouching stand-alone or plug-in software , it is usually because waht the SOFTWARE decides to do with any given subject. The programs I have tried out and decided to use routinely in my studio are the ones with many sliders that enable the operator to modify and control each kind of correction. There is however a caveat, the operator should have a good understanding of facial structure as well as the basic principles of portrait retouching.

I'll explain. Basic skin softening and removal of blemishes is easy enough, a simple healing tool can do that and the degree of softening is just a matter of taste or the client's requirements. When, however, you begin to alter facial structures, change the appearance of the eyes, reshape the mouth, whiten eyes and teeth, slim down a face, or remove detail for the orbital areas of the face (under the eyes), you gotta really know what you are doing. Portrait retouching is an artform, skill and talent in an of to itself. Nowadays, in digital retouching, you don't need as much manual skill as in the old method with pencils, dyes and etching knives, but you still need insight into facial structure and lighting.

So...I am not what I would call an expert master retoucher but as a portrait photographer, I wanted to learn as much as I can to maximize my knowledge, enable myself to do some basic retouching, and be able to communicate and direct retouchers that work in my studio. I took in several seminars given by well-respected master retouchers and although their styles, tastes and techniques varied, there was one important principle that they all agreed upon, that is you have to know WHAT to retouch WHAT NOT to retouch, and WEHN TO STOP RETOUCHING. Basically, if you remove certain lines, natural contours, and indentations on the face, the results will resemble a mask, If you seriously over-retouch a face you will end up with what they jokingly called "an embalming job"!

If you want to work "freehand" with frequency separation, other than simple blemish removal, besides the low and high frequency and layers techniques, you better understand facial structures and norms, otherwise, you are better off with the preset programmes and just learn-to control the sliders to your taste and/or the client's expectations. My full-time in-house retoucher has her own ways and uses frequency separation. Me- I am a big fan of Portrait-Pro but there there are also caveats which I will explain.

If you are the photographer AND the retoucher you have full control over how you light, pose and shoot each subject so you can create files that do not require radical or complex retouching techniques and make certain decisions, at the camera, to decide whether you are going to remedy a problem on the shoot or in retouching and make the best compromises. For example- let's say the subject has a serious double chin or jowls around the chin-jaw area. A slightly higher camera position and higher lighting ratio and more shadow around that area might solve the issue. Suppose, however, the subject also has a very long nose that will intersect with an upper lip at a high angle- you can't shorten the nose effectively in retouching but a lower camera position will fix that so you religate the chin/jaw issue to retouching. There are many other facial charismatics such as asymmetrical facial structures, uneven eye sizes, rough skin and much more that can be addressed in shooting lighting and just tweaked in post-processing and retouching.

The job is to NOT make the subject unrecognizable or perform radical "plastic surgery" but rather to deemphasize certain aspects of the face and body that the subject my consider undesirable or distracting.

There are subjets and times when NO retouchg is required or when everyoneagrees that "just let it all hang out"! There is nothiing wrong woth a detailed portraot study with all the "warts", however, I am in the BUSIBNESS of portait photography and that is not always up to me. Most folks don't mind a little beautification and some demand allot!
When photographers like or dislike any particular ... (show quote)


E.L. Shapiro,

Thank you for an excellent discussion and pointers I expect many will appreciate. The ON1 Portrait AI does a lot of things automatically if you desire and it has a wealth of sliders to allow the retoucher to tweak appropriately. Understanding lighting and camera angles and body positioning is better than trying to fix a bad shot in post IMHO. I started doing portraits back in the 1960s with film of course and you had to get these factors right. I was fortunate that I luck out and became the photographer of a modeling school and of a number of Playboy Bunnies who needed portfolios to get a better job. Now I'm retired and often take photos of people at ballroom dance events we attend. Most of the ladies are 60+ and love my "magic camera" as they say because they think they look ten years younger in my photos. I am supporting your point that changes should be not to morph the person, but to subtly enhance the photograph so that it still looks like the subject and strengthens the strong/good features and weakens the detracting features. Wrinkles are interesting in that they just need to be slightly softened, not removed. Hmmmm.... I wonder how those models and Bunnies look today.
Go to
Sep 25, 2020 21:50:45   #
ON1 has released their Portrait AI ahead of ON1 Photo RAW 2021. Very impressive so far.
Go to
Sep 19, 2020 14:47:20   #
rmalarz wrote:
I just happened upon a presentation by Blake Rudis. For those of you unfamiliar with Blake, he owns f/64 Academy and presents techniques and practices that enhance processing photographs. This presentation focused on the use of AI.

His contention is that its use is up to the individual but there are consequences. Aren't there always?

AI is a new buzzword. There's always going to be the next best thing. There is always something to entice the average photographer to purchase new software. There will always be improvements otherwise there's no need to upgrade software. But, is AI the way of the future, Blake doesn't think so. He then presents 3 good points why.

1. AI is just a buzzword. It's nothing more than a buzzword to make folks think this is the future.

2. Absolutely nothing can replace handwork and experience. If one wants to be a true artist, one is going to want to spend the time making the best one can make. That instead of letting some artificial intelligence take over and do the work for you.

If one is jumping on the AI bandwagon, they are probably, at best, an average photographer. Any true artist-photographer is going to put in the hard work and personal experience to obtain the best photographs they can create. Artificial Intelligence just can't do that.

3. When one places their trust in AI what they are saying is they are letting some company determine the benchmark for their success.

Trying something and succeeding provides motivation to keep going. Trying something and failing leads to learning and betterment. Without the bedrock foundation of failure, one can't measure success.

Until AI can express the emotional experience of producing a beautiful photograph, it's not something on which to rely. It simply allows one to think they are a better photographer then reality will show.

This brings up the speed at which processing can be done. Sure the one-button approach to processing leads to more time of going out and clicking that shutter. I agree with Blake. Photography involves both making the exposure and making a photograph. If one is going to rely on AI to produce that photograph, one will never be any better than the software they use.

These are some of the points Blake made in his presentation. I agree with them. It might be the dividing line between being a photographer and being a button pusher.

If you're interested in Blake's presentation you can search for

Blake Rudis 3 Reasons why AI is NOT the future of Photo Processing!
--Bob
I just happened upon a presentation by Blake Rudis... (show quote)


Hi Bob,

I must disagree with Rudis that AI is a buzzword. It is a field of study in computer science that I have been involved in for over three decades. AI is the top tier designation and comprises many subareas such as expert systems, machine learning, deep learning, neural networks, and so on. Can AI technology be used to replace humans doing certain work. You bet it can. Robotic manufacture of automobiles and quality control activities for example. Another, we is still unnerving to me, is AI being used to implement self-driving trucks which are currently being tested on the open highways in same places. As the story goes, back in the late 1960s, one of the major airlines said they had implemented a fully computerized commercial airplane. On its initial flight, the system came on the intercom and announced to the passengers as the plane was starting its roll down the runway that the airplane was under full computer control and no crew was now needed. All passengers need not be afraid of this new technology because Nothing can go wrong, go wrong, go wrong, ...

For photography, AI can be a great aid to us in post-processing, but the creative tweaking we all do will remain until the system can analyze enough of your work to learn your style and then you are no longer needed. What you think an image should look like and what I do may be similar or quite different.

Doc Barry
Go to
Sep 11, 2020 01:15:47   #
JD750 wrote:
👍

That is a nice shot. Good composition and warm colors.


Thank you JD750.
Go to
Sep 11, 2020 01:11:02   #
JD750 wrote:
Ok, so how do you explain the "golden hour" ? Shouldn't that be white? Perhaps the atmosphere has some effect on light color?


Yes, you are correct that the atmosphere affects the color. In the photo below I took on the Big Island, the Sun still appears basically white. The Sun is about three Sun diameters from starting to set or about 6 minutes (the movement of the Sun by one Sun diameter is 2 minutes since the angular diameter of the Sun is 0.5 degree). As the Sun begins it setting for the day, the color will change from a whitish color to orange to reddish. This is due to the atmospheric refraction that causes the blue and green light to be refracted more than the orange and red light. In addition, this refraction causes a foreshortening of the Sun. You may know the old sailor's rhyme of "Red sky at night, sailors' delight; red sky in morning, sailors take warning." The red sky is more dramatic if there is a bit of dust in the air.


(Download)
Go to
Sep 10, 2020 16:06:39   #
oregonfrank wrote:
When I took this image last evening in the context of heavy forest fire smoke the sun, to my eye, was a deep red/orange in color. But my D850 recorded the sun's color as nearly white (see image). Why the discrepancy?


Likely because the sun saturated the sensor. Look at you histogram.
Go to
Check out People Photography section of our forum.
Sep 4, 2020 17:58:41   #
11bravo wrote:
My understanding is that's a "club", similar to ON1's preferred membership, but the basic software is still "buy and own". Could be wrong...


Thank is my understanding too. I have Skylum, Topaz, and ON1 products plus PS, LR, etc. ON1 has an add-on club so to speak that lets you sync your work to all of your devices. I didn't sign up for that since I don't care to sync my work between devices. Business models continue to change to allow these small companies to stay in business. The big boys forced the cloud-based programs on us, so now the smaller ones can do it too or a hybrid of the model.
Go to
Sep 3, 2020 22:48:28   #
bsprague wrote:
"Topaz Mask AI is clearly deep learning based."

That would be fun reading. Can you send me to a link? Thanks.


Hi John,

See https://topazlabs.com/mask-ai/. They say machine learning which is the uber level of deep learning. The description of how they developed the program I read a bit ago certainly appeared they used deep learning principles in developing their algorithms.
Go to
Sep 3, 2020 22:12:41   #
JohnSwanda wrote:
Isn't Photoshop's Content Aware Fill considered AI?


John,
PS's CAF I believe arguably can be considered AI-based; however, Topaz Mask AI is clearly deep learning based.
Go to
Sep 3, 2020 18:28:00   #
bsprague wrote:
I wonder why Topaz and Luminar use AI? Not sure what photographic AI is, but why doesn't Adobe use AI?


Hi,
Photographic AI in this context is used to minimize your efforts in obtaining quality post-processed images. In theory, you would just input your photograph and the program would analyze it for content and then figure out what it can do to make it a fantastic final image. What the program thinks is fantastic may not be what you do, but then you have the option to tweak it as you desire. Actually, calling it AI is a bit too broad. In this case I expect Luminar AI actually is machine-learning based with deep learning methods (a subset of machine learning) to develop its "intelligence." Most likely Skylum has taken untold numbers of images to analyze (both good and bad) in the building of its knowledge base. Topaz has taken a similar path although they have been building AI modules that can interface with each other. Also they have been consolidating some modules. Over the next few years, I anticipate that Topaz will offer an fully integrated package with Topaz Studio 4 (my new number for the product!) as the base. Skylum and Topaz both make good products and are actively developing them.

Why doesn't Adobe have AI-based products for photography? Adobe does offer a product that is AI with machine learning, viz., Sensei which is for business applications. My answer to your question is that photography users are a relatively small fraction of their total business. Many businesses use Photoshop too I note. So Adobe wisely focuses their resources on products that affect the bottom line the greatest. Sensei makes sense for them. On the other hand, photography is the prime business of Skylum and Topaz Labs. Moving into AI is no small feat and it gives them both an edge in the marketplace. I note also that ON1 is also now incorporating AI in some features of their program ... this is also their business focus.
Go to
Page: <<prev 1 ... 5 6 7 8 9 10 11 12 13 ... 48 next>>
Check out Bridge Camera Show Case section of our forum.
UglyHedgehog.com - Forum
Copyright 2011-2024 Ugly Hedgehog, Inc.