Camera thread for marketers

  • Thread starter Thread starter Guest
  • Start date Start date
A narrow slice of a subdivided industry. I see what actually gets sold and in what volume. Just because there aren't forums devoted to stuff like PTZ cameras and studio-style production equipment doesn't mean that stuff isn't big. Just because most people here don't get all excited about a camcorder with built-in zoom doesn't mean that such products are outselling other sectors of the market, because they are. A single multicam reality show can be interesting and I enjoy posting about how Deadliest Catch switched to 7 VariCams and 20 EVA1s last season (up for an EMMY for it too! ;-D), but those 27 cameras were a nice sale but not a grand overall trend in the vast industry in which we all play. There are more thing in heaven and earth, Horatio, than are dreamt of in your philosophy.

Ok, i suppose we should specify that the we are not talking about broadcast pros, installs, and pure eng.

“Pro” here is more of the shooter that aspire to some form of “cinema”, whether it be a feature film, or a docu style promo shot in a verite’ style. They value getting through the day, making compelling content for a profit.

“Consumer” is the hobbyist and the prosumer. They want something that looks good and is affordable enough, since there will be no ROI. They may want the best and are affluent enough to buy whatever they want, or they may be trying to shoot snow floakes with a diy macro lens on a t2i. Who knows, but they care about image quality, as they are not content with a cell phone or PnS.

There are plenty of overlap between the two, but camera design can be divded into the two segments. What i gather from forums, is that people who work in the industry are often less demanding when it comes to image quality. Mainly because they have a lot more control over set wardrobe subject and light.
 
Last edited:
A narrow slice of a subdivided industry. I see what actually gets sold and in what volume.

It is a pity that those figures are likely to be confidential/commercially sensitive as I am sure they would make fascinating reading on the state of the whole market and it's many sub divisions.
Besides I suspect a huge part of it is down to what people want (as opposed to need) and the way the marketing plays into that perceived need.

3D came and went. Do we really need 6,8, 12K? "everyone" is now watching on 4inch phones. (Ok.... and a 65inch TV at home) that said my 4K 55inch TV upscales 1080p very well.
I note on Still cameras the professional bodies aimed ant the Press are 16-18Mb Then there are the ones aimed at studio use at 30-40Mb. The cameras in between 20-35Mb are aimed at consumers who still think they need more Mp to look at pictures on line or on prints at A4.

FWIW I have seen PTZ cameras mounted on a booth doing interviews at a trade show. They had 4 of them all driven by 2 people in the back who were doing production. They said it made a lot of sense on space and not needing 4 operators. It seems 1 operator per camera is long gone.

Incidentally the main evening news for the BBC has a studio with 6 remotely controlled cameras on tracks. A lot of the moments are pre-programed.
 
I note on Still cameras the professional bodies aimed ant the Press are 16-18Mb Then there are the ones aimed at studio use at 30-40Mb. The cameras in between 20-35Mb are aimed at consumers who still think they need more Mp to look at pictures on line or on prints at A4.

This is also reflected in much of what we read on dvxuser, in regards to video specs. Pros tend to be on the conservative side on one end, and then a few pros are on the edge of what is technically possible, staying ahead of a curve. But hobbyists tend to be a bit better than the average pro in some regards. My list of specs above for pros and consumers is taken from this forum.
 
Would the pro that prefers 8 bit to 10 bit please raise their hand?

There was the whole 8bit vs 10bit, and the verdict was: most screens are 6-8bit, and it is even uncertain if the data coming off the sensor is higher than 8bit, so 8bit is good enough, as long as it is done properly. Plus it is more effecient. 10bit is a waste, and might as well shoot raw if you want “quality”.

With the recent prolification of HDR, 10bit H.265, and rec2020, perhaps 8bit is now outdated.
 
It is a pity that those figures are likely to be confidential/commercially sensitive as I am sure they would make fascinating reading on the state of the whole market and it's many sub divisions...
I eyballed ARRI totals in a different thread - ~ $400M in revenues. Let's say 3/4 of it is camera based and an average ARRI goes for $60,000, then they ship about 5,000 units worldwide. Give or take.
 
There are a lot of video users beyond the ones i am listing, but listing the infinite types of customers is beyond this. For example. There is a whole “airsoft” genre of shooter. The type of people that buy something affordable that makes them feel like working professionals, even if they aren’t. Much in the way that people will often play army, and go buy airsoft guns that look real, and train and have competitions etc. If Red wasn’t actually some of the best cameras out there, i’d say that is what Res users are. But red has perhaps just tapped into that as a brand.

But i think moat brands tap into that “airsoft” aspect. Sigma FP is toting the line. Black Magic does it too with some of their gear. Could almost be argued that the Panny S series is semi airsoft. “Look how cheap a camera that is almost as good as xyz is!” and yet has few real pro features, and lacks the image quality a hobbyist might want.
 
I eyballed ARRI totals in a different thread - ~ $400M in revenues. Let's say 3/4 of it is camera based and an average ARRI goes for $60,000, then they ship about 5,000 units worldwide. Give or take.


Arri makes a lot of different products and rentals. How did you get the 3/4 number?
 
Saw this specs wishlist shirt, though they could be actual specs for a future camera.
48623805967_7953cbd319_z.jpg


It reads:
4K FF 8192x4320
240fps
12bit
4:4:4
Dyanmic range 16.5
Prores XQ 444
Prores raw
 
Last edited:
and it is even uncertain if the data coming off the sensor is higher than 8bit,

The "data" coming off the sensor photosites is analog voltage, so it is all up to the next step of A to D conversion. Do you sample the bucket in 8 slices, 10, 12? The higher the dynamic range, the taller the bucket, and that really needs more slices to define what's in the bucket. A high DR 12 bit camera may have slices that are as big and a low DR 8 bit camera.

When we get into 32bit float sampling on those A/D converters, then we will really be somewhere. Even if the rest of the pipeline is just 12 bits, that 32bit float on the front end can be used by a high power DSP to do a lot of stuff for down stream use. Not sure how much power that would take with current technology, but I'm guessing cooling needs are beyond what we can currently do.
 
The "data" coming off the sensor photosites is analog voltage, so it is all up to the next step of A to D conversion. Do you sample the bucket in 8 slices, 10, 12? The higher the dynamic range, the taller the bucket, and that really needs more slices to define what's in the bucket. A high DR 12 bit camera may have slices that are as big and a low DR 8 bit camera.

When we get into 32bit float sampling on those A/D converters, then we will really be somewhere. Even if the rest of the pipeline is just 12 bits, that 32bit float on the front end can be used by a high power DSP to do a lot of stuff for down stream use. Not sure how much power that would take with current technology, but I'm guessing cooling needs are beyond what we can currently do.
Why did you write data in quotes? It’s not like it isn’t information. The sensor is made to spit out usable information, otherwise it is a really bad solar panel.
 
Last edited:
The "data" coming off the sensor photosites is analog voltage, so it is all up to the next step of A to D conversion. Do you sample the bucket in 8 slices, 10, 12? The higher the dynamic range, the taller the bucket, and that really needs more slices to define what's in the bucket. A high DR 12 bit camera may have slices that are as big and a low DR 8 bit camera.

When we get into 32bit float sampling on those A/D converters, then we will really be somewhere. Even if the rest of the pipeline is just 12 bits, that 32bit float on the front end can be used by a high power DSP to do a lot of stuff for down stream use. Not sure how much power that would take with current technology, but I'm guessing cooling needs are beyond what we can currently do.
I’ve disliked 8bit since cameras could handle more than 10stops of dynamic range. Was back when the C300 and 1DC were introduced that all the 8bit debates were in full. I probably could have just written 10bit for the pro camera too, but the list is intended to be on the conservative side, plus i hadn’t seen the 8 vs 10 bit debate revisited yet. I am guessing the 8bit pros are now extinct.

It will be pretty awesome once we start getting into 16bits and more than 15 stops dr. Maybe even into the 20+ stops range. Of course, monitors will really have to up the game. It is kind of sad they didn’t move to 10bit before 4K, but at least 10bit is trickling in. 8K video on 8bit screens is really opening up the uncanny valley.
 
Last edited:
Some people forget just how much Arri does in the way of lighting.

The high end lighting market is far more competitive than the high end camera market, both in terms of sources and brand names.

Arri makes a lot of different products and rentals. How did you get the 3/4 number?
Out of my butt.

But, even if every dollar of revenues was camera based, it'd still be around 7,000. So, subtract lenses, lighting, licenses, Alexa65 rentals, etc. 5,000 is a good round number.

Oh and few years ago, Mitch asked how many Alexas I thought were out there. I speculated about 50,000. He said that was way too high.
 
Last edited:
The high end lighting market is far more competitive than the high end camera market, both in terms of sources and brand names.

Out of my butt.

But, even if every dollar of revenues was camera based, it'd still be around 7,000. So, subtract lenses, lighting, licenses, Alexa65 rentals, etc. 5,000 is a good round number.

Oh and few years ago, Mitch asked how many Alexas I thought were out there. I speculated about 50,000. He said that was way too high.

If there were 50K of them out there... we’d see a lot more crap Alexa footage, and ebay would be full of them.
 
Back
Top