Julian

Great work.

 

I said it before, but I'm gonna say it again: the G6 is the most underrated camera of 2013...

Probably we will see a refresh this year, wonder what the G7 will bring!

Most probably an issue with the adapter.

The for sale threads in the anamorphic forum keep it quite lively and bring about discussion of various unusual lenses. If we hid these away in a sub-section it might make the place a bit more dead.

 

What do you think?

 

Most for sale topics in the anamorphic forums have 0 discussion (or even replies). I have the feeling a lot of people just come here trying to offload their crappy anamorphic lens for an inflated price, or just drop a link to ebay.

 

'?do=embed' frameborder='0' data-embedContent>>
'?do=embed' frameborder='0' data-embedContent>>
'?do=embed' frameborder='0' data-embedContent>>
'?do=embed' frameborder='0' data-embedContent>>
 

Other suggestion: Selling topics allowed within the forum, only if you post proper footage (self made, not some random vimeo cat test).

I suspect though, that faults would get less and less, due to experience.

Faults are just part of the process. I don't know why, but this is always the case in chip production. Memory, cpu's, gpu's, etc.

Very interesting.

 

I wonder why they insist on making the wafers circular.

 

36% of the area wasted on the full frame sensor yield is pretty wasteful.

 

It's not really a matter of insisting. It has to do with the process of making the Silicon wafers and there are numerous other reasons for it, apparently. Some interesting answers here.

 

For example:

 

Nathan Zeldes

January 22, 2007 at 10:28 am

A key process step that uses the round form of the wafer is photoresist application – this is done by spinning the wafer rapidly and dripping the liquid resist at the center; centrifugal force takes care of spreading it evenly.

 

 

Another issue is the size of the wafer. They have been using 200mm for ages now I think. There is some switching going on to 300mm and 450mm is in development I think, but it probably will take a decade or more before it is standard, because all the machines have to be replaced and remade for the bigger wafers.

 

Btw... imagine the cost of that new 50MP Sony medium format sensor :) The price of the Pentax 645Z is incredible if you take that into consideration.

 

/Edit:

Interesting stuff on wafer sizes here.

57.8% of chip production (not only image sensors!) is already done on 300mm wafers. This will be 70% by 2017, but 200mm wafers are still being used for image sensors...

Did I miss anyone? Julian would you like to have moderator powers too?

 

Thanks but no thanks Andrew :) I know I spend (too) much time here, but I can't and don't want to commit to anything right now.

 

I'd still suggest bringing back the 'marketplace' sub forum though. I'll gladly report al the offenders who still dump their ads in the general or anamorphic forums. With the new police it should be easy to move them swiftly :)

Good decision Andrew!

Certainly, I think that a mFT sensor might cost one quarter of the raw materials to make (864mm^2 v 225mm^2 surface area = 3.84 greater area for FF sensor). I assume that higher pixel density would increase the cost though on smaller sensors? IMO I would guesstimate the price difference between mFT and FF as being a factor of three times. The main issue though would be R&D. Followed by where the sensor lies on the "Learning Curve". The Learning Curve means that as time progresses, things get cheaper to produce.

 

The cost of producing big sensors rises exponentially. It doesn't have anything to do with pixel density. Smartphone sensors are dirt cheap and have very tiny pixels/high density.

 

Sensors are made out of round 200mm (8 inch) silicon wafers. The bigger the sensor, the more (expensive) material you loose. This image explains it well:

 

 

But more important is the yield. When producing sensors, you probably never have a 100% yield. I have no idea what the failure rates are, but the tiniest piece of dust or whatever can ruin a sensor. So if one fullframe sensor is affected, a big part of the wafer is lost.

 

I have no idea what the cost is of a wafer. Tried looking it up and found this, that gives a 'typical processed wafer value' of $11,000.

 

11,000/24 would be ~$458 per fullframe sensor IF the yield is 100%. Lets say there are two faults on the wafer, two sensors fail, now the cost is already 10% higher per fullframe sensor.

 

Compare the Nikon 1 sensors to this (I'm not going to calculate the rates for M43... be my guest :)). 11,000/244 = ~$45 per sensor. If this wafer has the same amount of faults you can still get 242 sensors out of it. The price increase will be less then 1% per sensor.

 

Kinda off topic, but I felt like giving some insight in the cost of a fullframe sensor. I do agree that the A7S doesn't have to be more expensive than the A7 seen from the 'cost of sensor' perspective.

The 18-35mm is just too popular to keep up demand for Sigma (because it is f*cking awesome!). That's not the same as postponing a release and not delivering on promised dates like BMD.

 

Anyway, I can't tell you when it will be shipping. Check some other dealers? Here some shops have it in stock.

 

Anyone with their eyes on the new 50mm.. better preorder it if you want it soon :)

Did you test maximum exposure time for an electronic shutter still? It was 1 second with the gh3. To have long exposure possible electronically would be a big deal for timelapse people. I've asked a few places and it seems nobody has looked into this yet.

Thanks.

 

It's still 1 second on the GH4 (firmware v1.0). Just checked.

The GH4 was announced way before NAB. I think it make sense for those sites to go after the 'newest' news.

Some short thoughts:

 

Frankly it is becoming a real pain to moderate and I just don't have time for it.

 

That's quite noticeable. I can understand it, but I think it's not good for the community. The cause is quite simple, the forum is too big for you (or anyone else) to manage it on their own.

 

I am considering a change to the forum which would effectively see it closed but for a few selected members who have a proven track record of good posts. The forum would have a lot of exposure and readers. Like a 'second blog' to EOSHD that everybody can get something interesting out of.

 

I'd strongly advice against that. You might find it a real pain in the ass, but I'm sure a lot of people value the information and the knowledge that is shared here even with all the bullshit and silly arguments attached.

 

Also, I don't see any need in closing things down. If you want to go this way, just make a new sub forum for people who have earned their marks and leave the rest as is. Let people throw mud as much as they want and don't bother with it. I think visitors don't have much trouble filtering out the bullshit and choosing who's opinion they value or not.

 

I agree that the anamorphic forum has become a mess. Personally the people that are just hanging around here to sell stuff annoy me the most. I think you should bring back the market place forum. When it was here, nobody used it, but with a zero tolerance policy people will have to. You just need some volunteers to moderate basic stuff like that.

Definitely not expensive. This Ursa isn't for me, but I like the hdmi version much more than the Panasonic UGUYAYAYHHAYUA for example. A smaller/cheaper version of this would actually be pretty nice for the dslr crowd. A complete external recorder, screen + audio solution + battery in a box for one price.

Whats your experience with rolling shutter on the GH4 Andrew?

Version 1.0? Very nice.

 

Yep. Preproduction body with firmware-update! I shot with firmware 0.5 a few weeks ago, not sure if anything changed though.

 

Did you figure out the 'best' settings yet? Should I leave sharpness/contrast to 0 or dial it down in the Cine-modes?

Get the Sigma 18-35mm f/1.8. If you don't have the budget for it, sell your primes ;)

By the way.... just got this in ;) on loan.. sadly.

 

Anyone else think this is an absurd way of presenting a test anyway?

 

Nah, it totally shows the point of the test: you can turn darkness into daylight.

ISO 102.400 looks workable even on YouTube, imagine with proper processing.

 

It's not really hard to find the answer if you just put your topic title in Google :)

http://forum.blackmagicdesign.com/viewtopic.php?f=2&t=15382

 

BMCC 2.5K RAW = 5MB/frame. 7.5GB/min for 25p
BMCC ProRes = 1.6GB/min for 25p

 

4K ProRes: 240Gb = 36 Mins

Here's a 16 bit PNG file, converted from a DPX file output by gh444:

 

'>

 

Can someone (Andrew?) grab the exact same frame from the 4K source file and post it as a 16 bit PNG file?

I see the point of this adapter (looks expensive though), but why the hell put a crappy Canon 18-135mm on a Panasonic... Blasphemy!

Indeed Pentax have already announced that their 645Z will record 4K video.

 

Source? I think that's just a rumour for now.

_If_ 4K 4:2:0 really turns into 1080p 4:4:4, then 1080p 4:2:0 makes 540p 4:4:4. So downscaling to 720p would result in something between 4:2:0 and 4:4:4 (4:2:2 I guess?).

 

I don't really see why you'd want to do so though, unless you are watching on a 720p screen it will be upscaled again to 1080p, introducing artefacts and degrading the image quality.

True. It surely isn't a video optimised sensor. Probably more in the line of the 36 MPIX Sony chip in the A7R/D800. But still, fun times!

According to PhotoRumors:

 

The new Pentax 645z medium format camera will shoot 4k video, announcement on April 14th

 

In addition to full HD (1920x1080 @ 60i and 30p in H.264), the upcoming Pentax 645z medium format camera will also have 4k video recording capabilities (3840 x 2160).

 

 

If true, wow!

 

It has the same Sony 44x33mm sensor as the Hasselblad H5D-50c, up to iso 6400.

Wonder what bitrate/format it will shoot...