Panasonic Lumix DC-G9 now $1,297.99

Just bought an open-box G9 from Samy's via eBay, $999 wit regular panasonic Warranty, eBay ad said they have 3 more open box available

 

Wait, where do you get 10 extra? Is that a promo somewhere?

NM: it's on the page itself that notes it. didn't see that.

 

I paid $1365 US on release for my G9, back in January this year. Haven't regretted one single second of it.
Paid $498 US for my brand new, local stock GX8 not long after it was released. Still have it too. I think I'm ahead  

 

US has the most ridiculous prices. Never seen anything like what you get over there in EU.

 

Speedy must be a Panasonic employee and gets the discount.

 

I actually appreciate the 'skimming model' as you describe it.

I don't think it creates consumerism at all. To the contrary, I think it slows down the natural market cycles. For those who absolutely gotta have the newest new thing, there's a higher price for them, which helps the companies recoup their R&D costs more quickly. For the rest of the product cycle - that can last years - there's a lower price that lets the same product attract a broader market. This means fewer new models need to be delivered at any given time to flesh out a complete product line-up. The "mid-range" model is just an older "high-end" model.

 

No. Introductory pricing combined with 20‰ off eBay sale for one, a crazy 2 day sale plus sales tax refund for the other Right place right time Pretty much all my gear is bought this way now. I'm patient Absolutely zero affiliation with Panasonic.

 

Wow, that is amazing.

 

I was utterly convinced I was not going to upgrade to the G9. Look how that turned out. I would have got another 10% off it, but the wife wouldn't claim the tax back for me when she travelled outside the country.
I've also managed to time sales & travel for a few of my lens purchases as well. Better in my pocket

 

I think it depends on the consumer. I see a lot of 20 somethings going absolutely off the depend when it comes to consumer electronic purchases.

 

Some suggestions that the GH5s would get AI enhanced auto-focus like that promised for the coming Panasonic full frame bodies.

Little reason it couldn't also be rolled out to the GH5 and G9. Training up a machine learning algorithm can massive amounts of processor time, but running the resulting algorithms can often be done on weak processors.

Panasonic does not tend to release firmware feature updates for products that are going to be replaced any time soon. The fact that they're releasing feature updates at all suggests there are no near term replacement plans for any of those bodies.

As for the GX85 and G85/G80, Panasonic should really consider granting them firmware feature updates as well if they're not going to be replacing them with newer models. It's not surprising that the replacement cadence of their M43 products are lengthening. Not only because of engineering distractions of the full frames, but because the ILC market is shrinking.

Much cheaper to keep an existing product relevant with some firmware updates than engineer an entirely new hardware design.

 

It has nothing to do with ff distractions or a shrinking market.

All the manufactures are adopting one of 2 strategies.

1. frequently release a new version that has minor improvements over the previous one (apple iphone like)
2. release new versions less frequently with more substantial updates

This mostly comes down to the fact that with regards to sensor tech they are well into the realm of diminishing returns. When some completely new tech like organic senses make it to market then updates will be more substantial and more often.

 

A shrinking market absolutely tends to reduce product cadences. And the ILC market is absolutely shrinking.

Engineering resources are a limited commodity. Adding more engineers does not tend to scale well.

Logically, many of the hardware and software engineers Panasonic has previously used for M43 have now had their talents re-assigned to full frame.

Disagree? Read the Mythical Man-Month. A blunt summation, nine women can't have a baby in 1 month.

 

I've read some Panasonic literature somewhere, that states that the G9 does already have self learning autofocus.

 

In general, machine learning is tremendously asymmetrical. Requiring large amounts of computer time to train, but once trained, produces code able to run quickly on minimal processors. Google has said that many of their image recognition processes required months to train on dedicated server farms (effective super computers). Yet the algorithm that training created is able to run locally on an average smart phone.

For instance, all the machine learning enhanced camera processes on Google's Pixel phones run locally on the phone, and in many cases do not use much processor. We know this because Google's camera application has been ported / hacked to run on lesser phones with much less processor capability.

 

Here ya go, from the horses mouth
Fast, skilful and sleek LUMIX G9: Delivering best-in class performance for the photo enthusiast – Panasonic Australia News

And I quote from that link
Contrast AF with DFD (Depth From Defocus) technology has been further advanced, with the G9 able to achieve an industry-leading AF of 0.04 secvi (approx.). The camera’s high processing performance also allows it to incorporate Deep Learning technology (in addition to the conventional Face/Eye Recognition) – which detects the human body and is intended for action and sports photography.

 

In general, self-learning is not machine learning.

To achieve optimal A.I. enhanced focus, Panasonic would provide a training algorithm with large data sets of example imagery. Scenes that are in focus, out of focus, and most importantly, scenes that tend to confuse their current implementation of auto focus.

The algorithm created by this process would then be tested against similar scenes, but none it has actually experienced before. If the algorithm performs well on the previously unseen data, it could then be rolled out to production. If it does not perform well, then more data and more training would be required.

The user's use of that algorithm can never improve it. If users discover edge cases that the algorithm performs poorly upon, Panasonic would need to recreate those conditions in their facility, using those edge case scenes to further train the algorithm. After testing, that revised algorithm could then be sent out as a firmware update to users.

This could be what they're doing to improve GH5 and G9 auto-focus.

 

Still available.
https://www.adorama.com/ipcg9a.html?emailprice=t&utm_source=rflaid67913