Sony marketing holds down quality innovation for us pro users?

mmm said:
Steve, when I've been referring to 50Mbps recording in this thread, I have been referring to the XDCAM HD codec which is CBR, 50Mbps, 8 bit, 4:2:2 (which I believe requires 2 encoding chips). So to be clear, I think Sony should have planned on offering an option to record 50Mbps 4:2:2 XDCAM HD to the onboard SXS cards for the reasons we have both given, i.e. it makes life much easier for broadcast shooters, for whom XDCAM HD is "good enough."

Sony have their 35Mps 8bit 4:2:0 VBR MPEG2 codec which they call XDCAM EX and their 50Mbps 8bit 4:2:2 CBR MPEG2 codec which they call XDCAM HD. I think my abbreviating this to "50Mbps" led you to think I meant a new 4:2:0 50Mbps codec. I don't think adding such a codec would be a good idea.

This option would neither cost nor hinder ANY other user. It would only allow Sony to take some of the customers who will now buy a Nanoflash or Ki Pro (etc) in order to meet broadcast specs.
Click to expand...

Within the context of this thread, that 'marketing' not a technical issue is 'holding down' the quality of this camera. Raising the existing codec from 35 mbps to 50 mbps is a non-issue with cost, which is why I assumed that is what you are referring to.

However, adding a new dual-chip encoder is a completely different cost issue. The camera itself is priced that way it is because it can use the existing EX as a base. It doesn't need any dramatic re-engineering. Production can take place in parallel to the an existing model without manufacturing bespoke components.

Keep in mind, changing one component means a cascade of other components needs to be changed as well. Then you have to think about supply-chain and inventory of the new components, ip and licencing, component suppliers, etc.

It would have added another $1-2k to the cost of the camera, versus the $1-2k you would give to Convergent or AJA. But by building the cost into the machine, you force others that don't need it to pay for those features. By externalizing and modularizing the consumer can save that money and spend that money where they want.
 
Had it not been for the fact that the PMW-500 was recently launched with the ability to record XDCAM HD422 (MPEG2 50Mb/s 4:2:2) I'd have made the suggestion that I did about the choice to use AVCHD on the AF100 - that they don't have a codec of higher quality available for the media... For Panasonic DVCPRO HD and AVC-Intra only exist on P2, so without adding P2 to the AF100 they couldn't go beyond AVCHD. Until recently the same was true of SxS and XDCAM EX for Sony - the highest quality you got on SxS was EX 35Mb/s - for anything higher you had to use Professional Disc. But then they went and muddied the waters with the PMW-500 and allowed XDCAM HD422 on SxS cards, so now it seems unfitting that EX 35Mb/s is the highest quality offered onboard with the F3.

I'd not heard that two chips were required to exceed EX-quality, but that's interesting and brings a technical element back into it. I did know that Convergent Design used the same Sony chipset in their NanoFlash that Sony used in the EX1 and EX3, so it certainly did seem like a very artificial restriction to limit to EX in the F3.

The problem is then that the F3, out of the box, is not "HD" quality according to the world's most influential broadcasters, whereas lifting the rate to 50Mb/s 4:2:2 would have made it totally acceptable as Full HD virtually everywhere. Obviously there are heaps of products that will help fix this, but recording to an external recorder, no matter how small and lightweight, is significantly more complicated (and annoying) than onboard recordings.

Perhaps BBC/Discovery/etc will revise the spec in light of the massive sensor and low noise in the F3, but it seems unlikely.
 
Barry_Green said:
You have no idea what is or isn't possible at this price point
Click to expand...

I do! The answer is RAW. If Red was doing it in a $15,000 body 4 years ago, and now in a $5000 Scarlet body, then there is absolutely no reason that we are not seeing RAW codecs in Panasonic and Sony cameras now. Especially since Red had less experience, resources, and manufacturing capabilities. The only difference is marketing. Red only had one camera. Panasonic and Sony have an entire line and price points to protect. They'll never give us RAW because they're always going to want to sell us the "next best thing" "or broadcast standard". These standards are how these companies stay in business. If Panasonic or Sony rolled out with a s35mm imaging block with RAW recording for $10-20k then they would never sell another camera again. That's it, quality is maxed out. Until people need a bigger format or silicon improves. RAW is just not profitable unless you are a boutique cinema company like Red. Even then... what happens when all the studios have a Red One? How will Red generate income now? Oh wait... you can see it happening now with the Epic... "4k? Ummm, that's basically for idiot DSLR shooters... You need 5k now!" So much for rendering obsolescence obsolete. I do love what they are doing though, but I can see now why Pannavision never made their cameras available for purchase.
 
Last edited:
Red did it for 15k on a barebones body that didnt include a damn thing, not even a viewfinder or way to power it, and they were still losing money. Also, the scarlet is not going to be anywhere near 5k, its going to be well north of 10. And once again, that is for a barebones body, no batteries or media or anything else.
 
Cbrianchester said:
I do! The answer is RAW. If Red was doing it in a $15,000 body 4 years ago, and now in a $5000 Scarlet body, then there is absolutely no reason that we are not seeing RAW codecs in Panasonic and Sony cameras now. Especially since Red had less experience, resources, and manufacturing capabilities. The only difference is marketing. Red only had one camera. Panasonic and Sony have an entire line and price points to protect. They'll never give us RAW because they're always going to want to sell us the "next best thing" "or broadcast standard". These standards are how these companies stay in business. If Panasonic or Sony rolled out with a s35mm imaging block with RAW recording for $10-20k then they would never sell another camera again. That's it, quality is maxed out. Until people need a bigger format or silicon improves. RAW is just not profitable unless you are a boutique cinema company like Red. Even then... what happens when all the studios have a Red One? How will Red generate income now? Oh wait... you can see it happening now with the Epic... "4k? Ummm, that's basically for idiot DSLR shooters... You need 5k now!" So much for rendering obsolescence obsolete. I do love what they are doing though, but I can see now why Pannavision never made their cameras available for purchase.
Click to expand...

Not everyone needs or even wants raw and the headaches that go with it.
 
DylanReeve said:
I'd not heard that two chips were required to exceed EX-quality, but that's interesting and brings a technical element back into it. I did know that Convergent Design used the same Sony chipset in their NanoFlash that Sony used in the EX1 and EX3, so it certainly did seem like a very artificial restriction to limit to EX in the F3.
Click to expand...

That is an incorrect statement. It is not the chips from the EX1/3 in the Nanoflash to give 4.2.2 at higher bit rates. Its from the PDW-700, a thirty thousand dollar camera.
 
Last edited:
Cbrianchester said:
I do! The answer is RAW. If Red was doing it in a $15,000 body 4 years ago, and now in a $5000 Scarlet body, then there is absolutely no reason that we are not seeing RAW codecs in Panasonic and Sony cameras now. Especially since Red had less experience, resources, and manufacturing capabilities. The only difference is marketing. Red only had one camera. Panasonic and Sony have an entire line and price points to protect. They'll never give us RAW because they're always going to want to sell us the "next best thing" "or broadcast standard". These standards are how these companies stay in business. If Panasonic or Sony rolled out with a s35mm imaging block with RAW recording for $10-20k then they would never sell another camera again. That's it, quality is maxed out. Until people need a bigger format or silicon improves. RAW is just not profitable unless you are a boutique cinema company like Red. Even then... what happens when all the studios have a Red One? How will Red generate income now? Oh wait... you can see it happening now with the Epic... "4k? Ummm, that's basically for idiot DSLR shooters... You need 5k now!" So much for rendering obsolescence obsolete. I do love what they are doing though, but I can see now why Pannavision never made their cameras available for purchase.
Click to expand...

It's not just about K of resolution, but the M-X was a revolution in noise floor as well.

And c'mon, obsolescence obsolete is a marketing moto. We'll see how much it translates into partial truth now with modularity.
 
Duke M. said:
That is an incorrect statement. It is not the chips from the EX1/3 in the Nanoflash to give 4.2.2 at higher bit rates. Its from the PDW-700, a thirty thousand dollar camera.
Click to expand...

I've heard from numerous sources, including a Sony engineer, that the Convergent Design recorders have the same Sony OEM ASIC that are utilised in the EX-series of cameras.
 
Steve Castle said:
It would have added another $1-2k to the cost of the camera, versus the $1-2k you would give to Convergent or AJA. But by building the cost into the machine, you force others that don't need it to pay for those features. By externalizing and modularizing the consumer can save that money and spend that money where they want.
Click to expand...

Steve, I think we are actually in agreement now that I clarified the 50Mbps ambiguity. The only thing I would add is that if I were Sony I would have added XDCAM HD as an onboard option in order to capture additional £/$ that will otherwise go to AJA and CD. Even at $1-2K, I think many documentary shooters would have opted for the convenience of such an upgrade instead of an external recorder.
 
to understand why each new camera is not rewriting history, you need to understand how the corporate guys are thinking.
They do not ask the tech guys to create a RED each time. They are just looking to release a new product that is "at level" with other brand,
but the first question they have to solve, is how to integrate that product in the line of the existing stream.
Because it is not only one product, it is many products packed into one.
The codec XDCAM is one, some proprietaty chip are , the memory card is another, the license they have paid are ones.
So by producing a new camera, the goal is to make these others product alive as long as possible and get the return on investment.
So they cannot get a new chip, a new codec , a new memory card each time they release a new camera, because they have
teams that are working on the existing base, with an existing knowledge they have to use.
And what a pro would say if by purchasing a new camera , he would be obliged to get rid of his existing costly XDCAM production equipement (disc reader, recorder, player, software)
just because the new camera is not compatible with the old format.
If you are a pro, the workflow and the equipment needed to follow the workflow is much more important than getting that little increase in bandwith.
 
DylanReeve said:
I've heard from numerous sources, including a Sony engineer, that the Convergent Design recorders have the same Sony OEM ASIC that are utilised in the EX-series of cameras.
Click to expand...

An ASIC is an application specific integrated circuit. Think of it as a GPU on a computer video card. Throw two GPUs on a video card and it takes a new circuit board, new operating system, new supporting chips, etc. Its a whole new chip set even though individual packets of transisters (ASICs) may be the same.

I'm sort of laughing at the Red company's stance. For years they derided Sony, Panny and Canon for 'crippling' cameras to get extra marketability out of a camera line. Now that Red is marketing a line of cameras they are doing the same thing with the Epic lite being a 'crippled' Epic.

No body has crippled anything. They are simply not reinventing the wheel each time.
 
nosys70 said:
to understand why each new camera is not rewriting history, you need to understand how the corporate guys are thinking.
They do not ask the tech guys to create a RED each time. They are just looking to release a new product that is "at level" with other brand,
but the first question they have to solve, is how to integrate that product in the line of the existing stream.
Because it is not only one product, it is many products packed into one.
The codec XDCAM is one, some proprietaty chip are , the memory card is another, the license they have paid are ones.
So by producing a new camera, the goal is to make these others product alive as long as possible and get the return on investment.
So they cannot get a new chip, a new codec , a new memory card each time they release a new camera, because they have
teams that are working on the existing base, with an existing knowledge they have to use.
And what a pro would say if by purchasing a new camera , he would be obliged to get rid of his existing costly XDCAM production equipement (disc reader, recorder, player, software)
just because the new camera is not compatible with the old format.
If you are a pro, the workflow and the equipment needed to follow the workflow is much more important than getting that little increase in bandwith.
Click to expand...

The point being Sony already has the ability to deliver a universally accepted HD recording on SxS cards - they do it in the PMW-500. It records XDCAM HD422 (50Mb/s 4:2:2 Long-GOP MPEG2) on to Sony's standard SxS cards.

By not allowing the F3 to also record the same codec they're essentially shooting a camera that is not suitable for HD acquisition (for many broadcasters) without the addition of an offboard recorder (probably from a third party). As well as the expense and possible increase in workflow complexity you get with an offboard recorder, you also now have another device to carry/mount, connect and power. And it offers no on-screen feedback about media utilisation or battery state, or connection integrity.

This is totally acceptable when we're talking about niche use situations like 4:4:4 and S-Log where very few users will need that functionality anyway. It's also acceptable when we're talking about a ~$6,000 "prosumer" level camera - but this is a different beast. It's a serious HD camera that's being positioned in Sony's CineAlta range and is being priced to a serious broadcast market.

I think it's a failing on Sony's part (and I am a big fan of the camera and even the EX codec) to overlook this. It makes the F3 a much harder sell. As a buyer I go "that's a lot of money for a camera that only shoots EX" - of course I'm told that I can simply attach a NanoFlash (or similar). Then my response is that I can attach a NanoFlash to an AF100 or a Super35 NXCAM as well and get similar results for a lot let money.

The F3 has a lot of great features and a lot of potential with the S-Log and Dual-link options - but that's just potential. On it's own, camera-for-camera with the AF100 or Sony's own NXCAM offering, it doesn't compete as well as it should for the premium price.

Duke M. said:
An ASIC is an application specific integrated circuit. Think of it as a GPU on a computer video card. Throw two GPUs on a video card and it takes a new circuit board, new operating system, new supporting chips, etc. Its a whole new chip set even though individual packets of transisters (ASICs) may be the same.

I'm sort of laughing at the Red company's stance. For years they derided Sony, Panny and Canon for 'crippling' cameras to get extra marketability out of a camera line. Now that Red is marketing a line of cameras they are doing the same thing with the Epic lite being a 'crippled' Epic.

No body has crippled anything. They are simply not reinventing the wheel each time.
Click to expand...

Well specifically I was actually told it was the same chipset, so more than just the ASICs, although they'd be the key component. I don't work for Sony or Convergent Design, so I can't know for sure - but that's been well known as a "fact" around my area for some time. Perhaps the complexity to go from 35/4:2:0 to 50/4:2:2 is enough to require an expanded circuit, I don't know. I'd been led to believe that the EX cameras were physically capable of more, but limited to the 35/4:2:0 they have. Which, given their pricing and position in the Sony broadcast family, makes sense and is reasonable.
 
Thank you for this interesting conversation. I wonder who is thinking to invest in this camera, and what will be your plan to make a profit with it. What place in the market will it have. Did you calculate the costs of a focuspuller. Primes. External recording. Or is it on your budgetproduction with the build-in codec and some old photolenses.

So.. How do you see your market with this little strange nice camera :)
 
DylanReeve said:
Perhaps the complexity to go from 35/4:2:0 to 50/4:2:2 is enough to require an expanded circuit, I don't know. I'd been led to believe that the EX cameras were physically capable of more, but limited to the 35/4:2:0 they have.
Click to expand...

The Nano has two chip sets instead of one though.

Can you imagine an EX1 hack to 50Mbps and 4.2.2 similar to the GH1 hack? Not too many people would be willing to brick their EX1 but fun to dream.

I think a lot of budget people will be waiting for the NXCam in the summer. The SNR will be a deciding factor for me. The F3 at 63db is pretty hot stuff. The EX1R, which is very clean when adjusted right, is 54db. AF100? The HVX is pretty noisy. Who knows? If the NXCam has the F3 sensor and 64db I won't mind using it with a Nano and I'll probably use the AVCHD as back up.
 
Last edited:
DylanReeve said:
The point being Sony already has the ability to deliver a universally accepted HD recording on SxS cards - they do it in the PMW-500. It records XDCAM HD422 (50Mb/s 4:2:2 Long-GOP MPEG2) on to Sony's standard SxS cards.

By not allowing the F3 to also record the same codec they're essentially shooting a camera that is not suitable for HD acquisition (for many broadcasters) without the addition of an offboard recorder (probably from a third party). As well as the expense and possible increase in workflow complexity you get with an offboard recorder, you also now have another device to carry/mount, connect and power. And it offers no on-screen feedback about media utilisation or battery state, or connection integrity.

This is totally acceptable when we're talking about niche use situations like 4:4:4 and S-Log where very few users will need that functionality anyway. It's also acceptable when we're talking about a ~$6,000 "prosumer" level camera - but this is a different beast. It's a serious HD camera that's being positioned in Sony's CineAlta range and is being priced to a serious broadcast market.

I think it's a failing on Sony's part (and I am a big fan of the camera and even the EX codec) to overlook this. It makes the F3 a much harder sell. As a buyer I go "that's a lot of money for a camera that only shoots EX" - of course I'm told that I can simply attach a NanoFlash (or similar). Then my response is that I can attach a NanoFlash to an AF100 or a Super35 NXCAM as well and get similar results for a lot let money.

The F3 has a lot of great features and a lot of potential with the S-Log and Dual-link options - but that's just potential. On it's own, camera-for-camera with the AF100 or Sony's own NXCAM offering, it doesn't compete as well as it should for the premium price.
Click to expand...


The PMW-500 is a $25k camera that is specifically targeted to a segment of the market. The fact is 50/4:2:2 may be 'good enough' for certain broadcast consumers, but its not good enough for many others. Others that may need 4:4:4, S-log, 880mbps. People that need more than 50/4:2:2 will ultimately need to get an external recorder either way.

I understand your position, but every camera doesn't have to be build to meet every single regional consumer need.

So the question becomes, should this F3 become a $16-20+k camera and have 50/4:2:2 built in, or should we make the F3 a $13-16k and just allow the consumer to add what they need. Or should we just go all out and have a 4:4:4 SR Memory built in to begin with, make it larger and redesign it from scratch, and make it a $30-40k+ camera?

Yes, an external recorder is less elegant than an internal one but its a cheap and quick way to get FAR beyond 50/4:2:2. Its been made clear from the very beginning that the SxS 35/4:2:0 is really designed for offline editing, since you can record simultaneously with the 4:4:4 S-log footage.


In this camera we have a ~$13k camera that uses a Alexa-like 3.43k sensor without line-skipping, does 12-stops DR with S-log, 4:4:4, 60p, dual HD-SDI, can be made into a 3D-rig via 3D link, PL-mount, actually affordable PL lens kits, and F3 mount for servo-zooms.

Hell, Sony is even coming out with a $6k NXCAM with the same sensor for those that only need 4:2:2. For the 'good enough' crowd that's probably the camera for them. The box-ish camera body even looks to have mounting brackets where accessories can be mounted on.
 
I'm not going to argue that every camera should be suitable for all users, but I do think that for the likely market for the F3 (rather than the AF100 or NXCAM thingy) the ability to record a widely acceptable HD acquisition format would make the camera much more valuable. An easier purchasing decision.

I'm pretty certain that whatever camera Panasonic brings out to challenge the F3 will offer an acceptable codec on board. It will come with AVC-Intra as standard (and AVC-I 50 as the "lower quality").

The added features and expandability of the F3 are great but probably only really valuable to a niche.

As a broadcast user 50/4:2:2 is definitely good enough (in fact the de-facto standard in my region). If I need to bolt an external recorder on to the camera to achieve that, and I don't have much need for the high-end (optional and costly) add-ons, then I'm going to look at the cheaper options that can do the same. I'd seriously consider an AF100 or a Super35 NXCAM instead, as they'll offer a very similar result when paired with an external recorder.

Were it to carry XDCAM HD422 onbaord then it would be a much more appealing option for many users.
 
You are perfectly right by the point you mentionned, but if the F3 was HD422 (50Mb/s 4:2:2 Long-GOP MPEG2) they would call it a new PMW-500 .
but if they sell it for the price of an F3, who would then purchase a F500 ?
So you are right, this is purely political decision to fill a market segment. If the camera does not fill you needs or expectation,
it is not sony or the camera that is wrong, it is you not looking in the right market segment.
 
Last edited:
The people who are buying PMW-500s aren't about to buy the F3 instead. They are very different cameras for very different users.

The F3, like it or not, is competing against the AF100 (until Panasonic comes out with a higher model) and Sony's own unlaunched NXCAM offering. The F3 has the ability to do a lot more than either of those cameras can, in the form of dual-link SDI, 4:4:4 and S-Log recording, but those are all options that have to be purchased. Out of the box it's a lot more like the AF100 and NXCAM-thingy than it might like - and when it's necessary to add an external recorder to get a widely acceptable HD acquisition format, then it becomes a less attractive option for some purchasers.

Perhaps then Sony could offer an XDCAM HD422 recording option in the same way that Dual-Link and S-Log will be options. I'm still not totally convinced that it's not something that couldn't be achieved will a firmware upgrade, but if not then it's quite likely that the camera is constructed in a modular enough way as to make it possible as a return-to-base upgrade.
 
DylanReeve said:
The F3, like it or not, is competing against the AF100
Click to expand...

The F3 was not meant as a competitor against the AF100. The camera was designed on its own, with no competitor of a lower level in mind. They saw what RED was doing, they saw a price point gap, they saw the DSLR craze, and they made their own camera to fill the spot that had a void. The AF100 is not even in the same ball park as the F3, the NXCAM or whatever its called is the camera to compete with the smaller DSLR/AF100.

The F3's sensor was built from the ground up for this camera, we are talking about a PL Mount native, 4:4:4, s35mm sensor camera... There is no comparison other than physical size of the camera body, to the AF100. The F3 is its own league and was never made as an answer to the AF100.

In fact theres no actual competition that could be made between the two. They are at completely different price points, with completely different limitations. I don't see anyone ACTUALLY having a hard time picking between the two. I see a lot of talk about this that and the other, but I find it hard to believe there is a large group of people whould actually be scratching their heads trying to figure out wether to spend $4,500 on a AVCcam limited 8bit hdsdi out, micro4/3's camera, or $13,300 on a 4:4:4, s35 sensor, S-log camera. They are just not in the same ball park. You're either perfectly fine at the $4,500 mark for a camera or 13,300 mark. not many would pick the AF100 OVER the F3 for features. Its purely a price point issue, if you are satisfied with the AF100 you're not at the same level as needing the F3 then.
 
Last edited:
You must log in or register to reply here.
Back
Top