CCD vs CMOS, again

[CMur12]

How about a diversion?

...

The next thing you know someone is going to try to prove that Keds are better in almost all cases than PF Flyers, and there is sure to be a big fight on the playground.

I preferred Keds, myself. (This makes you and me old!)

Thinking back, I think that Keds were more court shoes and PF Flyers were more running shoes. I would probably choose the latter, today (if I were a kid, once again).

- Murray

 

[David Hughes]

I will test and let the evidence tell its story. I programmed computers for a long time and we all held that one test was worth a thousand opinions.

I can't agree with this; taking a simple camera like the Digilux 2 or LC1 I can look at the menu and see 27 variations in "Picture Adjust" and then we can add 7 variations by bracketing in thirds from the so called correct exposure; making 27 x 7 variations (ie 189 variations). Now factor in handheld or on a tripod (making 189 x 2 = 378 variations. And there's flash on and flash off (thinking of fill-in adds another) so 378 x 2 = 756 variations without even taking the card out of the camera and printing it.

Of course, merely using the camera to produce prints might seem odd but I'm old fashioned. It might be best to say we all have our own opinion of what is best and better and leave it at that with everyone happy with their choice...

Regards, David

PS OTOH, if just one test of a computer is the standard it does explain a lot of things...

 

[Yokosuka Mike]

I can't agree with this; taking a simple camera like the Digilux 2 or LC1 I can look at the menu and see 27 variations in "Picture Adjust" and then we can add 7 variations by bracketing in thirds from the so called correct exposure; making 27 x 7 variations (ie 189 variations). Now factor in handheld or on a tripod (making 189 x 2 = 378 variations. And there's flash on and flash off (thinking of fill-in adds another) so 378 x 2 = 756 variations without even taking the card out of the camera and printing it.

Of course, merely using the camera to produce prints might seem odd but I'm old fashioned. It might be best to say we all have our own opinion of what is best and better and leave it at that with everyone happy with their choice...

Regards, David

PS OTOH, if just one test of a computer is the standard it does explain a lot of things...

David, I wish you would post more often. I always enjoy your take on things.

All the best,
Mike

 

[ptpdprinter]

How about a diversion?Thinking back, I think that Keds were more court shoes and PF Flyers were more running shoes. I would probably choose the latter, today (if I were a kid, once again).

The Keds vs. PF Flyers debate takes me back to elementary school, so let's say early 1960s. I remember being in the Red Rooster shoe store trying to decide which to buy. I think I usually got Keds but probably went rogue for PF Flyers a couple of times. You got all white Converse Jack Purcells if you were serious about tennis, but you bought those down at the tennis shop. I began running in the eighth grade and a new running shoe manufacturer named Tiger was taking the nascent running scene by storm. They are now called Asics. I ran cross-country in high school. The school actually bought shoes for the cross-country runners, but they cheaped out and bought us the bottom of the line Adidas. I believe the model was the Antelope, and they were the worst shoes I ever wore. You got blisters just looking at them. A couple of years later a little company called New Balance came along and mostly sold to long distance runners. Their shoes were lightweight and flat, i.e. no elevated heel. They were really comfortable but provided almost no cushioning. They had thin soles and wore out quickly. They were also expensive. Take the above with a grain of salt. Sometimes what I think are memories are just neurons misfiring.

 

[David Hughes]

David, I wish you would post more often. I always enjoy your take on things.

All the best,
Mike

Many thanks for that encouragement; I've not been able to post for a while as I've had some trouble with my eyesight that made using the computer difficult. The first bout of treatment finished earlier this week and I'm due back at the eye clinic soon (I hope) for the rest of it but I'm waiting for a lot of results from blood tests and an x-ray or two.

Thanks again.
Regards, David

 

[boojum]

<snip>

PS OTOH, if just one test of a computer is the standard it does explain a lot of things...

Yes it would explain a lot of things, wouldn't it? But after spending years writing code I can assure you that there is never "one test." Each and every possibility is tested as it must be and in many circumstances starting from the most basic line of code on up through the program as a whole and up into that program's system and on into the next larger system and so on up until it is included into the entire system. Large systems like the one I worked in took two or three days to run from end to end in batch. That was the first bill round of a huge billing system, the largest bill round.

Your example of 756 variations would have been tested many, many times because they can all happen in the real world and must be tested to see if they work and how they work. The adage of, "One test is worth a thousand opinions" is not a devaluation of testing but a dismissal of opinions as the bailiwick of fools in a scientific testing world. "Oh, that looks like it should work," is not worth the ink to write it. Ask anyone who has been there. There are a few on the board.

That computers work at all is nothing short of a miracle and is the result of endless hours of testing. To understand the concept try writing some code and getting it to work, something more complex than, "Hello world" and you will understand pretty quickly. When working with multiple variables within the program you are writing, testing will be revealed as immensely more helpful than "That looks like it should work" because it won't work. So one test with the little it reveals is immensely more instructive than the thousand opinions which are little more than the exercise of a fatuous ego. I could give you an example of a fool who changed an algorithm in a massive system upgrade which "looked like it should work." It took almost a week before the bug was identified and that meant a week's worth of work had to be rerun. That company is a cash business so it was borrowing money overnight to stay afloat, millions of dollars, and it took a week and a half to get caught up. Multiple millions were lost. The idiot who made the change got to leave with a glowing reference. They just wanted him out the door and gone and he was a contractor which made it even worse. Testing? Opinions? Yeah, tell me about it.

 

[David Hughes]

Boojum, I'm not going to disagree because I have been there and done that but as a user since the late 1950's or early 60's. Sometimes at the sharp end wondering why we bother when we had a manual version that worked and could be easily checked and corrected and, just before I retired a long time ago, sitting on a committee to try and define the UK's two penny worth of input to an international standard to get things like the internet working.

But sitting at home scowling at the laptop I often wonder if any testing is done and why they seem to go out of their way to complicate things. Thinking of all the ways you can (or could) exit a program in Windows is a good example and then there's all the different formats to save things in from say a word processor and so on. (I'd de quite happy to have stuck with "Word 97" but ...)

Regards, David

 

[boojum]

Boojum, I'm not going to disagree because I have been there and done that but as a user since the late 1950's or early 60's. Sometimes at the sharp end wondering why we bother when we had a manual version that worked and could be easily checked and corrected and, just before I retired a long time ago, sitting on a committee to try and define the UK's two penny worth of input to an international standard to get things like the internet working.

But sitting at home scowling at the laptop I often wonder if any testing is done and why they seem to go out of their way to complicate things. Thinking of all the ways you can (or could) exit a program in Windows is a good example and then there's all the different formats to save things in from say a word processor and so on. (I'd de quite happy to have stuck with "Word 97" but ...)

Regards, David

Not all software is the same. MS has a reputation for being more concerned with marketing. Market pressures cause apps to be released way before they are ready. I can only speak to my experience as a systems analyst which entailed rigorous testing. A major release could be in progress for over a year with all systems meeting monthly for progress reports. Each group which sent a rep had their own area of responsibility in which the code was written, tested at program level, internal test with test data, internal test with production data and finally running parallel to production runs with output compared, every data file and every byte of data within the file. When the system is billing the public there are a lot of regulatory agencies watching. Other shops I worked in had similarly high standards. It was a no jive, no BS world with very high standards.

What is peddled for PC's is a different world of cowboys racing to be the first in the marketplace. The customer is not the same as with internal systems where there is one customer who will fire your ass if you screw up too bad and too often. Too solve the MS problem I abandoned that world for Linux where things work better and are cheaper or free, more likely free.

And while we are at it, how many folks read the camera manual? ;o) And today's cameras, computers with lenses, are immensely complicated. So I run them pretty much as the come out of the box and at full auto. And you know what? My batting average is way higher with the auto cameras than it is with the manually operated RF cameras. Computers are smarter than I. They get the exposure, focus, white balance and image type correct and hand me back a nice photo. I'd never get all that stuff right and not having to do it allows me to compose the image. The RF cameras are an interesting anachronism. But as with all things here and elsewhere, YMMV.

 

[David Hughes]

Almost agree with you about the cowboys but the software I'm moaning about mainly is from old established firms; most of it started in Windows 3 and was posted to me on a floppy disc. They are still going strong but seem market driven, as you say.

I do the same with cameras; run them for a week on Auto everything except ISO and then give a mild tweak. As you say it seems to work with a few exceptions. On my pension I can't afford anything new and having been down that path don't want to; so your experience is probably different.

Regards, David.

 

[raydm6]

Unfortunately, with software products or software-based hardware (digital cameras for example), we are sometimes forced to abstract and learn the information architecture of a product or process to learn their affordances [SUP]1\[/SUP]. It can be quite numbing.

As you may know, there's a whole science devoted to the information architecture/usability/psychology/human-factors fields and you sometimes wonder how many resources or how much thought is devoted to this aspect of design. I think for some products; a lot. For others; not so much.

Interesting paper on affordances in design - (PDF):

AFFORDANCES IN PRODUCT ARCHITECTURE: LINKING TECHNICAL FUNCTIONS AND USERS’ TASKS
https://www.id.iit.edu/wp-content/up...0DTM-84525.pdf

With a long career in hi-tech (hardware/software Configuration Management and Product Lifecycle Management field), I am actually quite impressed with the current state of some software products (particularly cloud-based). All my work is cloud-based and currently can be done from my laptop from any location where my phone can be used as a hot-spot. No need for Wi-Fi (if not available) and my phone provides ample bandwidth for data uploads/downloads: never any latency issues. This is a wonderful way of working not available in the not-so-distant past. My phone acts as a back-up and except for screen real-estate, I can work from the phone.

In the past, we had to rely on tool-smiths/programmers to write APIs (application programming interfaces) to automate tasks and tie disparate systems together and to create workflows. Now, this function is mostly embedded in the design and APIs relegated to passing information between systems (for the most part).

In the past, I went from a DEC dumb-terminal connected to a mainframe running MRP software with manual paper-processes, to UNIX-based workstations: X-Windows/MOTIF, SUN SPARC-stations, then onto IBM-based PCs, MACs and lord knows how many secondary apps and software tools.

One had to be facile and adaptable to survive in hi-tech environments and since I interact with all aspect of the corporation (and supply chain), one cannot afford to be technology averse.

I understand how frustrating technology can be for some people (I have an older friend who uses a flip phone; can't use a smart phone) and I wonder if some camera makers should think of simplifying the user-menus or at least, offer varying levels of complexity for the user community.

--------------------------
1.\ https://uxmag.com/articles/affordanc...product-design

 

[DaveSee]

Unfortunately, with software products or software-based hardware (digital cameras for example), we are sometimes forced to abstract and learn the information architecture of a product or process to learn their affordances [SUP]1\[/SUP]. It can be quite numbing.

As you may know, there's a whole science devoted to the information architecture/usability/psychology/human-factors fields and you sometimes wonder how many resources or how much thought is devoted to this aspect of design. I think for some products; a lot. For others; not so much.

Interesting paper on affordances in design - (PDF):

AFFORDANCES IN PRODUCT ARCHITECTURE: LINKING TECHNICAL FUNCTIONS AND USERS’ TASKS
https://www.id.iit.edu/wp-content/up...0DTM-84525.pdf

With a long career in hi-tech (hardware/software Configuration Management and Product Lifecycle Management field), I am actually quite impressed with the current state of some software products (particularly cloud-based). All my work is cloud-based and currently can be done from my laptop from any location where my phone can be used as a hot-spot. No need for Wi-Fi (if not available) and my phone provides ample bandwidth for data uploads/downloads: never any latency issues. This is a wonderful way of working not available in the not-so-distant past. My phone acts as a back-up and except for screen real-estate, I can work from the phone.

In the past, we had to rely on tool-smiths/programmers to write APIs (application programming interfaces) to automate tasks and tie disparate systems together and to create workflows. Now, this function is mostly embedded in the design and APIs relegated to passing information between systems (for the most part).

In the past, I went from a DEC dumb-terminal connected to a mainframe running MRP software with manual paper-processes, to UNIX-based workstations: X-Windows/MOTIF, SUN SPARC-stations, then onto IBM-based PCs, MACs and lord knows how many secondary apps and software tools.

One had to be facile and adaptable to survive in hi-tech environments and since I interact with all aspect of the corporation (and supply chain), one cannot afford to be technology averse.

I understand how frustrating technology can be for some people (I have an older friend who uses a flip phone; can't use a smart phone) and I wonder if some camera makers should think of simplifying the user-menus or at least, offer varying levels of complexity for the user community.

--------------------------
1.\ https://uxmag.com/articles/affordanc...product-design

Greets, with hopes you & yours are, & remain healthy,
This, as has been posted, is a, um, tried debate regarding CCD vs CMOS sensor/detector. As we know, all digital sensors inherently record only luminescence, light...including mobiles & other digital sensors, regardless CCD or CMOS (CCD mobile?). All other factors are affected by sensor filters, as in Anti-Aliasing (AA, Leica missed this one with the M8, or did they?), & Bayer mosaicing implementation.
& then there's DOS(Disk Operating System) on a rig's processing chip: DOS is huge, yet it's not DOS, but PIC(Position Independent Code, 8 bit) which likely the functions...again, hmm, "which sensor best for PIC?" Oh!, & the on-chip OS may also be Linux, or BSD(Apple)... dunno
Thus; is it truly the sensor, or how implemented?
(Disclaimer: I've both CCD & CMOS rigs, RF w/primes & appreciate both.)

 

[Sonnar Brian]

The name PIC initially referred to "Programmable Interface Controller", before Microchip and soon after as Peripheral Interface Controller.
Position Independent Code usually means the code is relocatable in memory. This is a generic term, not specific to the Microchip PIC line of products.
Most 8-bit Microchip PIC code is Absolute. The PIC18 instruction set introduced relative jumps and calls, the later PIC14EXT instruction set adds them. The PIC12 and PIC14 instruction set uses absolute addressing for GOTO and CALL. You can perform math on the Program Counter for relative jumps such as "ADDWF PCL,F" where the W (working) register is loaded with the number of addresses to skip over. It works like a computed GOTO.

The Leica M8, M9, and M Monochrom cameras used the Renesas M16C processor and Analog Devices Black Fin DSP. An 8-bit PIC would be much too slow for the processing. I have about 20 years experience programming those. 40+ years of using DOS, the last 30 for embedded systems. It is fast, and gets out of my way when I want. I use PharLap TNT for a full 4GBytes for DOS.

 

[Godfrey]

I must be the odd man out on this topic of "reading the manual" ... I generally don't want to buy a camera or computer until I've already read the manual from cover to cover and think the device might suffice for my needs. 😉

I was also in high tech (mostly computer stuff) for a long, long time (32+ years) and made my living as a mechanic (and a photographer!) before, during, and after those years as well. No one builds something entirely without testing, testing, and re-testing until at least in some sense the thing works. No one does a shoot for pay without testing the equipment, understanding what needs to be done and how to do it ... no one who is successful, at any rate. Certainly no one who's income depends on it.

The fact that there are bugs in everything we make, one way or the other, simply shows that the things we have constructed are complex beyond the conception of how many different ways they can break or not work the way in the manner we intended. That's all. 🙂

G

 

[ptpdprinter]

I must be the odd man out on this topic of "reading the manual" ... I generally don't want to buy a camera or computer until I've already read the manual from cover to cover and think the device might suffice for my needs. 😉

I didn't read the manual before I bought my digital camera. I read a review. It sounded like it did everything I needed it to do. It took photographs. When I got it, I read the manual. Maybe I changed a couple of things in the menus. The only one I remember was setting up the parameters of Auto-ISO. That's a pretty cool feature. Otherwise, I just use my digital camera like my film camera. You know, take photographs.

 

[Sonnar Brian]

No one builds something entirely without testing, testing, and re-testing until at least in some sense the thing works.G

My Daughter has been hired to test the embedded systems I do. No one on the Planet is more skillful on finding mistakes that Dad makes. And the Bugs found get jumped on quickly!

 

[David Hughes]

I've lost count of the number of dud cameras I've bought - often for the opening bid - just to get the manual and - now and then - a pleasant surprise. Some of them seem impossible to get even as pdf's.

Regards, David

 

[Godfrey]

I've lost count of the number of dud cameras I've bought - often for the opening bid - just to get the manual and - now and then - a pleasant surprise. Some of them seem impossible to get even as pdf's.

Regards, David

That's odd. I've managed to find the factory instruction manual for every camera I've purchased since 2000 or so online first, whether the camera was new or old. And that's a fairly large number of cameras when I think back on it ... almost embarrassingly large. :angel:

G

 

[Godfrey]

I didn't read the manual before I bought my digital camera. I read a review. It sounded like it did everything I needed it to do. It took photographs. When I got it, I read the manual. Maybe I changed a couple of things in the menus. The only one I remember was setting up the parameters of Auto-ISO. That's a pretty cool feature. Otherwise, I just use my digital camera like my film camera. You know, take photographs.

Hah. I don't read reviews very often, other than for the entertainment value. I prefer to form my own opinions...

G

 

[ptpdprinter]

Hah. I don't read reviews very often, other than for the entertainment value. I prefer to form my own opinions...

I don't read reviews very often either, just when I am thinking about buying a new camera, which isn't very often. I have had two new cameras in the last ten years. I bought my current camera in 2016. I am not really very interested in a new camera at the moment. My camera works fine, and isn't holding my photography back. It does everything I need it to do, which is make photographs. And I prefer to form my own opinions too. You can form opinions in a variety of ways. Certainly reading the manual before you buy a camera as you do is one way, but not the only way. Buying a camera with a return privilege and trying it out is a good way too. I know you did that with the Pixii II, and returned it when the sensor didn't play well with some of you lenses. So that worked out pretty well. You seem to be doing okay, so I would stick with what you're doing.

 

[Godfrey]

I don't read reviews very often either, just when I am thinking about buying a new camera, which isn't very often. I have had two new cameras in the last ten years. I bought my current camera in 2016. I am not really very interested in a new camera at the moment. My camera works fine, and isn't holding my photography back. It does everything I need it to do, which is make photographs. And I prefer to form my own opinions too. You can form opinions in a variety of ways. Certainly reading the manual before you buy a camera as you do is one way, but not the only way. Buying a camera with a return privilege and trying it out is a good way too. I know you did that with the Pixii II, and returned it when the sensor didn't play well with some of you lenses. So that worked out pretty well. You seem to be doing okay, so I would stick with what you're doing.

🙂 Thank you for your approval. It is so appreciated. 😉

I often buy cameras (and other things) with a return option for the notion of testing them for usability in lieu of being able to get enough information otherwise. After reading the manual, of course, to determine if I'm interested enough at all...

G