Joel Halbert

Joel Halbert

Lives in United States Tucson, AZ, United States
Works as a Engineer
Joined on Jul 4, 2003


Total: 25, showing: 1 – 20
« First‹ Previous12Next ›Last »
In reply to:

maxnimo: Why do they call it a 1" sensor when not even the diagonal is even close to 1" in length?

It's a good question. Image sensor size is described by an arcane standard, and it goes like this:

Back in the early decades of early video, when imaging was done via electron (vacuum) tubes, the image was focused onto a photo-charge-sensitive plate. Typically, the cylindrical glass tube that housed the imaging plate had a diameter of about 1.5 times the imaging plate diagonal. Engineers would refer to the imager size by citing the tube diameter, and since most of the development was in the USA, the measurements were cited in inches.

So, a "one inch" imaging tube housed an imaging plate roughly 2/3 of an inch diagonal, while a "four thirds inch" tube housed an imaging plate just under an inch diagonal.

This nomenclature, as silly as at seems now, has stuck and is used for describing the size of many digital still-camera image formats.

Direct link | Posted on Apr 19, 2015 at 02:16 UTC
In reply to:

thejohnnerparty: Hum. 20 MP. If Sony can make 20 MP on the 1", why isn't the M4/3 going to be somewhere near 24 MP or 27 MP? That would get closer to the density of the 1". Hum?

The 20MP of the 1" sensor is not necessarily there for quality reasons, more for marketing. The total area of the so-called 1" sensor is just over half that of MFT.

If your belief is that pixel density = quality (quite the opposite of what some others would say), then MFT would need 36MP to compete. And APS-C would need 60MP. And FF would need 140MP.

Yet the achievable maximum quality, by most measures, increases with sensor area. Compact cameras, from very small-sensor ones up through 1" sensor and beyond, tend to market the MP number aggressively, limited more by practical storage and frame-processing limits than by real optical needs.

Direct link | Posted on Apr 18, 2015 at 19:15 UTC
In reply to:

vesa1tahti: Can somebody tell me what kind of sensor is inside the Lumix DMC-GX7. I just purchased the body for only 299 € (in Finland). I use Oly lenses like 14-42mm/3,5-5,6, 45 mm/1,8, and 12-40mm/2,8. The camera is impressive, my previous model Oly E-PM2 is quite excellent, too. Thanks!

GX7 was the first model to use Panasonic's latest-design 16MP sensor, with overall performance very comparable to, some say slightly better than, the Sony chip that was used in the Olympus E-M5 and the Panasonic GH3.

Since then, this GX7 sensor or a very similar one has been used in many cameras including the flagship cameras: Panasonic GH4, GM1, GM5, GF7, LX100 (non-interchangeable compact) and Olympus E-M1.

By the way, as part of various re-structuring activities to deal with the profit crisis, Panasonic partnered with TowerJazz who acquired 51% of the semiconductor fab operations. Essentially this was a way to shift losses and more effectively market the fab capacity to non-Panasonic customers (as Sony has for years).

Most people are assuming that Olympus will jump to a new Sony sensor in the next MFT generation, but this remains to be seen.

Direct link | Posted on Apr 18, 2015 at 18:44 UTC
On Panasonic Lumix DMC-GM5 First Impressions Review preview (301 comments in total)
In reply to:

NZ Scott: I just bought this camera after a three-week professional photoshoot in Australia and am really enjoying it. Certainly the controls are a bit fiddly, but with its tiny kit lens it feels almost weightless next to my E-P3 with 12-35 f/2.8.

I'm really looking forward to giving it a good nudge in the Philippines during the next two weeks.

Get ready, there is a giant storm heading towards the Philippines -best of luck with your travels and your photos.

Direct link | Posted on Dec 4, 2014 at 15:31 UTC
In reply to:

Serious Sam: The high ISO doesn't look too bad......but consider the X-trans can do just as good or even better. I am not so sure.

Best to wait for the DPR proper review and compare under the noise SD chart.

ajberg: I certainly did compare it to both. Sam's message didn't specify which Fuji, so most of my comments compared it to a camera that is more of a competitor in size & zoom setup. However, I did write that the APS-C Fuji is somewhat better, as I would expect from APS-C + prime lens.

Sam, I'm not defending nor commenting on the price, I'was genuinely interested in whether the LX100 is as much a failure as people were saying. I discovered it's not, but the jury is still out about the JPEG issues until we understand what is happening with the settings and the camera's attempts to deal with dynamic range - are we looking at over-ambitious iAuto processing?

And Peter - you might LOL but others here have done exactly that (compared it to 1/1.7") and claimed the LX100 is worse. So I've looked and posted about 1/1.7" LX7, 2/3" Fuji, 1" RX100, "sub-MFT" LX100, APS-C Fuji and FF Sony. Guess what, the results come in the expected order despite bashing comments. LOL doesn't change it.

Direct link | Posted on Oct 7, 2014 at 15:13 UTC
In reply to:

Serious Sam: The high ISO doesn't look too bad......but consider the X-trans can do just as good or even better. I am not so sure.

Best to wait for the DPR proper review and compare under the noise SD chart.

More on the Fuji: compared to LX7 with its smaller sensor than X20, the ISO 6400 shots clearly have more detail on the LX7, which I didn't expect. I think it's clear that the Fuji is using heavier NR; you can see the calculator looks "plastic clean" and the scales are unreadable in the X20 shot, readable scales but somewhat grainy surface in the LX7 shot. So, I conclude that Fuji sweetens the X-Trans magic with heavier default NR - one would have to compare RAW and/or play with both (X20 & LX7) cameras' settings to reveal true differences. LX100 is way better.

N.B. I am not claiming that the controversial LX100 sample shots here don't have any JPEG problems (honestly haven't decided), just that various sweeping claims about lens, sensor, noise etc. being "awful", "worse than my phone" are demonstrably over-the-top. Even "soft lens" is not supportable.

If I had pre-ordered (I didn't) I wouldn't be cancelling just yet. Objective tests look very good, full-iAuto _may_ be problematic.

Direct link | Posted on Oct 7, 2014 at 06:45 UTC
In reply to:

Serious Sam: The high ISO doesn't look too bad......but consider the X-trans can do just as good or even better. I am not so sure.

Best to wait for the DPR proper review and compare under the noise SD chart.

I was intrigued by your claim and did some Imaging Resource comparisons.

See my message earlier, may be hidden in the reply list to BozillaNZ. The comparometers are quite easy to use.

Anyway, the LX100 absolutely blows away the Fuji X20 2/3" at ISO 6400, not close at all. Basically the LX100 is quite usable at 6400 and you won't regret the shot; with the X20 (same as X30 in my understanding) you should definitely stay away from 6400.

OTOH, the Fuji X100 APS-C does somewhat better than LX100 at 6400.

Every comparison I have done on the IR site shows basically expected results in rank order, predictably by sensor size first, expected lens quality sort of second (complex issue) and sensor resolution last. LX100 is impressive and better than the cameras it should beat, not as good as larger sensors w/prime lenses that should beat it in this objective evaluation.

However, I can believe that the JPEG algorithm needs work especially when evening out sun/shadow lighting variations.

Direct link | Posted on Oct 7, 2014 at 06:19 UTC
In reply to:

BozillaNZ: I've got a developed raw image of the studio shot from Image Resources. This file is converted to Canon CR2 format using my own developed tool and then converted to JPG using, hold on to your seat, the Canon RAW Image Task. The color rendition is a lot more Canon-like and the sharpness is very good.

LX-100 SOOC:

LX-100 to Canon CR2, developed in RAW Image Task:

Thanks very much for doing this work. I gave in to the pixel-peeping and compared your conversion of the Imaging Resource shot to the same scene from Sony RX100-III and for a reference standard, Sony A7R with (apparently) 35/2.8 Sonnar. With the white yarn near the upper-right The LX100 appears a little sharper and reveals slightly more texture than the RX100-III, despite the claims here of lens softness and despite the resolution advantage of the little Sony. They are both more than satisfying. As expected, the A7R with prime is rock solid and unquestionably sharper.

Also checked LX7 based on claims here that it out-performs the LX100 (sorry but quite the opposite) and GX7.

So, no surprises:
1. Sony A7R + 35 Sonnar
2. Lumix GX7 + Oly Ft 50mm
3. Lumix LX100
4. Sony RX100-III (not massively below LX100)
5. Lumix LX7

Center results are in the same order though a bit better than the near-corner region (the IR scene does not have any content in the extreme corners). All ISO 200.

Direct link | Posted on Oct 7, 2014 at 05:28 UTC
In reply to:

FRANCISCO ARAGAO: More LX-100 photos:
"Gee, look at the wavy lines in the concrete wall - must be some kind of crazy lens distortion - software correction FAIL."

Only kidding - thanks for the links - no they probably aren't the right display format for pixel-level peeping, but they are very nice photos.

Direct link | Posted on Oct 5, 2014 at 19:31 UTC
In reply to:

cgarrard: Funny all the comments about the image quality. The greatness of this camera hasn't been nor will it be its amazing sensor quality. It's about the form factor surrounding the chip, and its lens, etc. Maybe Sony's 20mp 1" chip would have been a better choice for all of you complaining? Or even better, maybe Panasonic should have gone back to a 1 1/6.5" size chip instead and called it the LX9.


In this section you have two comments about "maybe they should have used the Sony 1" sensor", and this has been mentioned by others here. The LX100 sensor, even with the additional crop, is excellent.It is the same as in the GH4, E-M1 and other recent MFT cameras. The Sony 1" is nice but it is not as magical as some seem to think, and it is 20MP for marketing reasons. The RX100 lens is also nice, and small, but has significant limits as realists have noted. Panasonic could have (and may yet) use the 1" chip in a compact zoom as they have done with the FZ1000 super-zoom.

They chose to create a somewhat larger camera with a faster lens and flagship-camera sensor; this was not done to play any kind of prank on their customers. Once the dust settles, more results are published and the emotional pendulum quiets, anyone who doesn't have an anti-brand bias will come away with a more positive conclusion. Some will buy, some won't, but the camera is not some sub-smartphone joke.

Direct link | Posted on Oct 5, 2014 at 18:52 UTC
In reply to:

shutterbud: I agree these samples don't tell us much. I would have liked to have seen some sort of rhyme or reason to them. I think we need to remember we are dealing with 12 MP images of often distant objects. What is THAT all about? Does anyone considering this camera decide its best to use it to take shots of distant objects with small detail at wide angle? No! I think DPR should have a loose format within which to play around; brick wall at ten paces, portrait in sunlight, portrait in shade with fillflash, flowers.
We know what we want to see and we have not yet seen it; different skin-tones in difficult light, the limits of detail from the lens, considering it has only 12MP, micro-contrast from the lens. We need close-ups of dogs in sunlight, people of different races in mixed light, chessboards or wooden objects at hi ISO. Dynamic range challenges. Etc etc etc.
Think about it DPR.

Yes, if you look at various first-samples galleries you will see that the subject matter depends very much on who shot them, what the weather was like at the time, etc. Many galleries have the kinds of shots you mention, much less of the distant buildings and trees (the latter of which don't impress from any camera after a bit of enlargement).

Direct link | Posted on Oct 5, 2014 at 18:28 UTC

Manyt first-samples galleries on DPR and elsewhere produce disappointment in the comments. Looking at other samples galleries, natural textures (trees, bark, leaves, grass) tend to become quite indistinct upon modest magnification on-screen). Even the highest-end sensors and prime lenses produce mushed texture if you examine them critically. Sharp man-made edges (buildings, brickwork, slats) tend to fare better.

Regarding the lenses - looking at the Sony A7R gallery for example, a late FF sensor with Zeiss-brand lenses, I see a notable difference between the prime 35mm vs. the Vario 24-70 in terms of how well the detail holds up vs. enlargement. In this context, the LX100 samples, especially the .acr version, are not bad but if you want an extra enlargement factor of two to four more before you see mush, then you need to start with a Full-Frame sensor and a high-grade prime lens.

Direct link | Posted on Oct 4, 2014 at 22:15 UTC as 118th comment | 3 replies
In reply to:

Mattoid: Looks like some pretty extreme distortion correction is happening here, and not working. Im seeing weird wavey lines all over the place. Look at the green lampost in shot 4: its bent. As are the edges of the walls of the building in shot 8. ok its an old building but that bad? Shot 24, another example. Sometimes a bit of barrel distortion looks more natural. Did they just go over the top with correction or is there some weirder moustache like distortion caused by extreme light bending (to make the lens small) being hidden here?

I'm really trying to see what you are seeing. Many many of these samples are shot with a tilted camera at a fairly wide to very wide focal length, and so have perspective distortion. However, "bent or wavy" I'm not agreeing. I lined up a straightedge with each of the vertical posts in shot 4 and they seem fine. Shadows and subject context can affect the perception. I do not doubt that there may be aggressive un-bending going on in software, but it seems to have worked out successfully in my view.

Direct link | Posted on Oct 4, 2014 at 21:31 UTC
On Samsung introduces Galaxy Note 4 phablet with OIS post (75 comments in total)
In reply to:

Valiant Thor: As I observe people and society I am 100% convinced that the smartphone is the worst invention in the history of the world, bare NONE. It has turned a once observant and interactive society into pathetic, electronic addicted, anti-social zombies who walk around with their glazed empty eyes affixed to those crappy little LCD boxes glued to their hand, their thumb frantically pushing stupid little chicklets around the screen with the look of a meth addict on their face and mortal fear in their mind that they would accidently miss one non-essential, meaningless message from an emotionless fellow zombie that is doing the same thing. What used to be meaningful family social outings have turned into four people sitting at a table lifelessly and listlessly staring hypnotically at the endless supply of meaningless pixel propaganda which never leaves their hand. Other than that, nice phone!

Nice rant. By the way, it's "bar none" - perhaps your phone or tablet's auto-complete keyboard is to blame?

Fads always feed legitimate rants over lost traditions, talents and morality. Yes it is amusing to see a family gathered round a table with the smartphone glow on each face. I can't say that never happened in my family , but we have a non-strict understanding that mealtime is for conversation. Having said that, the phone can be used as a conversation starter or enhancer.

Every technology shift has its ups and downs. But the mobile internet also has tremendous value. The zombie-overuse syndrome is offset by its information and communication potential. Your "worst innovation" is also the single most important recent innovation for developing-world nations. In order to break the cycles of poverty, repression and low opportunity, folks have to have an observable alternative model. Even low-end smartphones hooked into a minimal communications infrastructure now provide that.

Direct link | Posted on Sep 6, 2014 at 19:25 UTC
On Samsung introduces Galaxy Note 4 phablet with OIS post (75 comments in total)
In reply to:

Jogger: I see a lot of women with Notes.. mostly because they are easy to tote around in purses/handbags. For men, its far too large unless you are using a man-purse or have extra large pockets.

I carry the Note 3 on my belt in a flip-cover hip pouch. Works quite well and makes the pocket issue moot. I do the same for pocketable cameras, which are even smaller and fit into a nice small belt pouch.

My wife keeps her Note 3 in her purse, of course.

Having become used to it, I wouldn't trade the Note for a smaller phone, and I don't think she would either.

Direct link | Posted on Sep 6, 2014 at 18:51 UTC
On Leica T (Typ 701) First Impressions Review preview (2251 comments in total)

I just posted a take on the "lie" accusation, if anyone cares to read it, in
though I realize it's par for the course to throw mud.

A common misconception, if you read these comments, is that software correction is a "cheat" that can make a bad lens good - not so at all - and that companies with high morals don't cheat, so a truly ethical firm will deliver an expensive multi-element lens instead of a cheaper, simpler lens plus software "cheats". If these people would just stop and think logically for a moment, perhaps they'd see the glaring flaws in this reasoning. See the referenced post for more.

I often wonder why these same people don't demand that their car engines should run well (or even run at all) if the engine management computer were disabled. It's part of the engineering solution, and in today's market it's not an option to do without, but it doesn't mean there are no justifications for costly engine builds vs. cheap ones.

Direct link | Posted on May 3, 2014 at 22:25 UTC as 181st comment | 3 replies
On Panasonic Lumix DMC-GX7 Review preview (599 comments in total)

First, I want to thank DPR and Jeff Keller for posting this review, despite the fact that "my camera" didn't quite get a Gold.

We all weigh pros and cons differently. For example, no Panasonic before (or Nikon or Canon or Fuji) has missed a Gold because it _has_ IBIS (in addition to most of its lenses having OIS) - but just not in preview or video mode(?!)

I do see most of the listed cons as minor - the pros may have been slightly under-appreciated.

Now specifically regarding the tilting EVF - it's a bit frustrating to read reviewers who can't fathom this feature. I find it to be a great help. Here is a thread where several of us addressed the topic:

Whys and Wherefores of a Tilting Viewfinder

Also, I find the EVF to be very relaxing to use at the height of a sunny day here in AZ, clearly more so than using the screen. I have family in Seattle though; from my travels I understand that the screen may typically be more usable there.

Direct link | Posted on Nov 7, 2013 at 06:25 UTC as 184th comment
On Panasonic DMC-GM1 preview (631 comments in total)
In reply to:

M0P03: First impressions preview didn't make clear this line: Image stabilization "Unknown".

Stabilization is in-lens OIS as with other Lumix models (excepting the GX7 which brings in-body stabilization for non-OIS lenses). You can see that the GM1 kit lens has OIS; most of the non-zoom lenses do not.

Direct link | Posted on Oct 29, 2013 at 06:38 UTC
On Panasonic DMC-GM1 preview (631 comments in total)
In reply to:

Joseph Mama: I wish they could expand some of their bizarre statements, such as "The Sony Cyber-shot RX100 offers more bells and whistles, but you're limited to its fixed zoom lens and much smaller sensor. "

What does that actually mean? How is a 1 sq inch sensor "much smaller" than a 1.66 sq inch sensor?
What exactly "bells and whistles" are on the RX100 that this thing lacks?

The sensor area ratio is closer to 2:1, not 1.66:1, and neither is close to 1 square inch: GM1 is ~0.35 sq inch, RX100 is ~0.18 sq inch. Whether you believe that qualifies as "much smaller" is of course a matter of opinion.

RX100 Mark II offers a tilting screen and a hotshoe/EVF mount, but plain RX100 has neither of those. I agree that the statements are not completely clear.

Direct link | Posted on Oct 29, 2013 at 06:34 UTC
In reply to:

Marty4650: Hasselblad could learn a LOT from Leica.

Hasselblad takes a nice camera, puts a clown suit on it, and sells it for a 600% premium. Then everyone has a good laugh.

Leica takes a nice camera, makes it even nicer without changing anything on the inside. They just improve the styling, create a bunch of nice cases and straps, and include some better software. Then they slap a more reasonable 80% premium on it. At least relatively speaking it is a lot more reasonable.

Then some people actually buy it.

You have to give Leica credit for this.

Agreed. By the way, the only $399 price I see is a used or refurbished one from Panacam. I am not aware that Amazon has lowered the LF1 price more than about 5% here and there - yet. Of course, lower prices always come to those who wait.

Again, the price delta seems surprisingly low for this one. We are talking about an evening or two out for many of the people here, or a good trip to the supermarket plus a gas tank fill-up. Or one extra night in a hotel on vacation. I'm not saying $200 is nothing, but in my book it doesn't paint someone as a filthy vile bourgeois fool idiot.

I think I'm talking myself into it just to raise some hackles around here. I like the burgundy one (I bought a red GF1 because I liked it, plus people here ridiculed it for not being "pro" black. But I do want a black GX7 since they don't make red).

Direct link | Posted on Sep 10, 2013 at 03:53 UTC
Total: 25, showing: 1 – 20
« First‹ Previous12Next ›Last »